immigrants in johannesburg: estimating numbers and assessing impacts

CENTRE FOR DEVELOPMENT AND ENTERPRISEInforming South African Policy

IMMIGRANTS IN JOHANNESBURGEstimating numbers and assessing impacts

IMMIGRANTS IN JOHANNESBURGEstimating numbers and assessing impacts


August 2008

CDE In Depth no 9

Published in August 2008 by

The Centre for Development and Enterprise

Transmedit Place, 5 Eton Road, Parktown, Johannesburg 2193, South Africa

PO Box 1936, Johannesburg 2000, South Africa

Tel +27 11 482 5140 • Fax +27 11 482 5089

[email protected] • www.cde.org.za

© The Centre for Development and Enterprise

All rights reserved. This publication may not be reproduced, stored, or transmitted without

the express permission of the copyright holder. It may be quoted and short extracts used,

provided the source is fully acknowledged.

ISBN: 978-0-9802628-9-6

CDE In Depth provides South African decision-makers with detailed analyses, based on original research, of key national policy issues.

Series editor: Ann Bernstein

This is an edited and abridged version of a comprehensive report based on

survey research on immigrants in Johannesburg. The full-length report, which

is available from CDE, was written by Professor Lawrence Schlemmer, who also

designed and carried out the surveys. This version was written by Dr Sandy

Johnston.

The Johannesburg Survey is part of a larger CDE project on immigration,

skills, and xenophobia, which aims to provide a research base for effective

policies to manage immigration. The project is led by CDE’s executive director,

Ann Bernstein, and managed by Dr Sandy Johnston. This project was funded by

the Atlantic Philanthropies. The funders do not necessarily agree with the views

expressed in this publication.

Contents

Executive summary 5

Introduction 15

Towards an estimate of the number of foreigners in Johannesburg 21

Assessing impacts: the balance sheet as seen by residents,

immigrants and officials 33

Summary of results 48

Concluding remarks 49

Endnotes 54

Most immigrants, at all levels – from the street trader to the

boardroom – make a contribution to South Africa’s economy,

sometimes against daunting odds. We need to understand,

manage and capitalise on that contribution much better. But we

will only be able to do this if we are able to manage flows of

people across our borders more rationally and efficiently

5July 2008 |

This is one of the

largest and most

sophisticated studies

on immigration

to South Africa

undertaken thus far

Executive summary

THIS REPORT PRESENTS the key findings of a CDE survey of cross-border immi-

grants living in Johannesburg.

It is being released in the aftermath of a wave of murderous xenophobic attacks on

foreigners in the city and surrounding areas that has brought migration issues into

shocking focus, and moved the need for more effective migration management far

higher up the policy agenda.

South Africa has attracted large numbers of immigrants, particularly from poorer

neighbouring countries, because of superior economic opportunities and services.

Concerns about migration and immigration must seek to strike a balance between

humanitarian concerns and international obligations; the need to control borders;

citizenship and social integration; and skills needs and the economy. However, the

country lacks the information needed to achieve such a balance.

This is one of the largest and most sophisticated studies on immigration to South

Africa undertaken thus far. Besides establishing a far more accurate profile of

immigrants in Johannesburg, it examines how immigration issues intersect with

public policy challenges.

The studyCDE’s research was devised in conjunction with leading demographers to reach a

population that wishes to avoid being identified and enumerated. Building on les-

sons learnt from an earlier pilot study of Witbank, the research was aimed at esti-

mating the number of foreigners in Johannesburg, finding out more about them,

and investigating the attitudes of South African city dwellers towards them.

The core element was a household survey aimed at assessing the number of immi-

grants in the sample, and obtaining estimates of the numbers of immigrants in

neighbouring households. This was followed by interviews with immigrants them-

selves. Interviews were also conducted with 45 representatives of business organi-

sations and 32 police, local government, immigration, and welfare officials. The

field work was conducted in the second half of 2006.

CDE’s survey was designed to be broadly representative of all areas and types of

residence in Johannesburg. The attacks on foreigners in May 2008 were largely con-

fined to informal settlements bordering on major urban areas, which represent only

one area of immigrant concentration. As a result, public awareness of immigration

issues – and such research as has been conducted since the violence – have tended

to focus on these areas. This report offers a broader view of the numbers and impact

of immigrants, and attitudes towards them, in Johannesburg as a whole.

It is also important to bear in mind – again in the aftermath of the xenophobic

attacks – that the survey does not distinguish between irregular and legal immi-

grants. The violence has tended to focus attention on ‘illegal’ immigration,

6 | Centre for Development and Enterprise

Executive summary

CDE’s survey suggests

that the number

of foreigners in

Johannesburg in 2006

was around 550 000

although much of the media and many members of the public have a poor under-

standing of the categories of legality involved in migration. This has encouraged

popular assumptions that most foreigners are criminals merely by being here. In

our survey, the views of and about those immigrants who are here legally are also

reflected in the findings.

How many foreigners in Johannesburg?Johannesburg is South Africa’s largest metropolitan area, and its biggest focal

point for cross-border migrants. This is because, first, it is the economic centre of

Gauteng, which is the motor of the national economy; second, it is in close interac-

tion with two other metropolitan complexes, Tshwane and Ekurhuleni; and third,

it is within easy reach of neighbouring countries Botswana, Mozambique and

Zimbabwe.

Estimates of the foreign-born population in Johannesburg and in the country as

a whole have fluctuated widely for several reasons: the main one is the need of

unauthorised migrants to avoid being identified. Then there is the difficulty of

distinguishing between migrants who have settled here and those who regularly

return home; the ability of migrants to illegally obtain documentation from cor-

rupt officials; and the expense of large-scale population research. Official figures

on foreign residents are captured in national censuses. This means that the most

recent official source dates from 2001, when Statistics South Africa estimated the

number of foreigners in the country as a whole at 2,3 per cent of the population (or

1,1 million people).

In the face of the difficulties involved in counting immigrants – especially irregu-

lar ones – wildly varying estimates are the norm. For instance, in 2006 a Johan-

nesburg-based economic consulting firm estimated that there were 9,84 million

irregular immigrants in South Africa at the time (an unlikely 20 per cent of South

Africa’s total population).

Analysis of CDE’s Survey data suggests that the number of foreigners in Johan-

nesburg in 2006 was around 500 000 to 550 000. Johannesburg’s population was

estimated by Statistics South Africa’s 2007 Community Survey to be almost 3,9 mil-

lion. Estimating a percentage for these figures depends on the degree of visibility

of (especially irregular) foreigners to the Community Survey. The upper limit of

CDE’s estimate of foreigners (550 000) would give the following range of percent-

ages for foreign residents in Johannesburg: if all foreigners are visible (and hence

the Community Survey total of 3,9 million is accurate) about 14,5 per cent of the

population of the Johannesburg metropolitan area is foreign. Depending on how

invisible foreigners are, the true population of the city could be as high as 4,45

million, and the percentage of foreigners as low as 12,35 per cent. Neither extreme

(complete visibility and complete invisibility) seems likely, so the figure is prob-

ably somewhere between the two percentages.

7July 2008 |

Executive summary

Xenophobia appears

to be inconsistently

related to economic

threat, education

and level of contact

Immigration as seen by Johannesburg residentsZimbabweans and Mozambicans are perceived to be the largest groups, followed

by Nigerians, Chinese, and Malawians; at a lower level are Indians, Pakistanis and

Europeans. Immigrants from other countries in Africa have low visibility because

of their small numbers, although Somalis have low numbers but high visibility.

Those from BLNS probably exceed Nigerians but have low visibility.

The dominant foreign groups (Zimbabweans, Mozambicans and Nigerians) are

seen to cluster in inner city and associated areas, although they also have a strong

presence in outlying areas. The composition of immigrants diversifies with move-

ment into the fringe areas, while the former white suburbs have a more selective

immigrant presence.

Foreigners are perceived to work as hawkers, artisans, miners, domestics and gar-

deners, shopkeepers and stallholders, in professional and technical activities, per-

sonal services, other services, security, and small manufacturing and crafts.

Negative perceptions of foreigners outweigh positive ones among black South

Africans, less so among minorities, although levels of intolerance appear to be no

worse than in other countries.

Xenophobia appears to be inconsistently related to economic threat, education

and level of contact. However, one pervasive factor is the myth that a foreign pres-

ence is tantamount to a criminal presence. The stress induced by high crime levels

creates a perpetual need for scapegoats.

Nonetheless, there is agreement among both blacks and minorities that certain

categories of immigrant are beneficial. BLNS citizens, Westerners and Indians

are highly approved; Chinese, Zimbabweans and Mozambicans approved; and

Malawians and Pakistanis approved with qualifications. Nigerians, Somalis and

people from non-English-speaking African countries (except Mozambique) do

not gain approval.

Characteristics of immigrantsThe results of the follow-up survey among 302 foreigners cannot give a complete

picture of the foreign residents of Johannesburg because the results probably

under-estimate the number of foreigners who live and work on suburban and

peri-urban properties, as well as former residents of the BLNS countries and immi-

grants who live in non-residential circumstances (sleeping rough, places of refuge,

workplaces). Nevertheless, the results are representative of those in the residential

fabric of the city, who are less reclusive.

The foreigners in our sample are concentrated in informal housing areas (41 per

cent) followed by formal townships (23 per cent), with the remainder evenly

divided between the inner city and suburbs.


Executive summary

The principal factor

determining perceived

levels of hostility is the

ability to blend into

local communities

Only some 20 per cent of the foreigners were unemployed. Forty-four per cent

were self-employed, and 12 per cent employed other people: the average number

of employees was four, half of whom were South African.

Their occupations were much the same as they were perceived by other residents,

covering a wide range of small trading and independent service provider occupa-

tions, as well as employment ranging from technical and artisan to professional.

Sixty per cent of foreigners experience negative treatment in Johannesburg. In line

with the various approval ratings, those from BLNS, Europeans and East Africans

experience less hostility than those from West Africa and Mozambique.

The principal factor determining perceived levels of hostility is the ability to blend

into local communities through cultural compatibility, closeness to one of South

Africa’s ethnic groups, and the ability to speak a South African language. Those

who bear the brunt of xenophobia are those who dress, look and speak differ-

ently, notably Nigerians and other West Africans, Somalis, and – to a lesser extent

– Mozambicans. When the fieldwork was carried out in 2006, the level of xenopho-

bia was tolerable for most immigrants, but some immigrant communities have for

years been targeted for harsher treatment and sporadic violent attacks.

Although 75 per cent of foreigners say that they intend to return to their countries

to retire, in reality far fewer actually do so.

Feedback from our survey interviewers indicated that they frequently suspected

that respondents were neither ‘illegal’ nor fully and legitimately legalised in terms

of formal entry requirements, but had acquired ‘genuine’ (in the sense of not being

forged) documents illegally through corrupt officials. Of course this could not be

openly said, but it is worth noting that experienced interviewers were strongly and

frequently given this impression.

Perceptions of local officialsCDE interviewed 32 high-level officials in national and metropolitan police serv-

ices, planning divisions of local government, welfare administration, and regional

immigration offices throughout the Johannesburg region.

Key points in officials’ responses were:

In-migration to Johannesburg is seen as out of control. Their (over)-estimates •

of the number of foreigners in the city average out at 2,5 million (in a population

officially estimated at 3,9 million in total), of which 86 per cent are believed to

be non-legal. This perception produces feelings of helplessness and despera-

tion among officials.

Only 12 per cent of them see immigrants as an advantage for Johannesburg, •

while 55 per cent see them as a liability.

Many officials concede that immigrants make a significant contribution to •

small business activities, with 71 per cent observing that foreigners mainly

9July 2008 |

Executive summary

Most officials

in the ‘frontline’

of immigration

issues believe that

Johannesburg’s

services and facilities

are seriously

oversubscribed

do jobs that locals do not. Two-thirds admit that foreigners are hardworking,

determined, and productive.

Most officials in the ‘frontline’ of immigration issues believe that Johannes-•

burg’s services and facilities are seriously oversubscribed and that foreign

migration is adding to the stress.

Some 50 per cent believe that foreigners are more involved in crime than locals; •

43 per cent believe them to be equally involved.

There was a broad consensus about the need for skilled immigrants in fields •

such as engineering, the professions, teaching, entrepreneurship, the con-

struction industry, artisan and technical trades, manufacturing, and IT.

Perceptions of business peopleSenior officials of all 45 significant business organisations operating in Gauteng,

including foreign chambers of commerce, were interviewed:

Some 49 per cent saw the large number of undocumented foreigners as a very •

serious problem, 27 per cent as fairly serious, 16 per cent as both a problem

and a benefit, and 9 per cent as a benefit rather than a problem.

On average, they perceived the number of immigrants in Johannesburg to be •

more than 2,7 million.

Twenty per cent felt that the contribution of immigrants to small businesses •

was considerable, and 44 per cent that it was helpful to some extent. Only 24

per cent felt that foreigners reduce business opportunities for locals, and 76

per cent felt that they add value.

Given skills shortages, some 30 per cent of the business organisations observed •

a great need to recruit foreigners, and 36 per cent a moderate need.

Many felt that South African unskilled labour has priced itself out of the •

market.

High crime levels were mentioned as the factor that most hindered the recruit-•

ment of skilled foreigners.

Summary of resultsThe number of foreigners in Johannesburg at the time of the study (2006) is esti-

mated to have been more than 500 000. Depending on their degree of visibility in

Statistics South Africa’s population estimates, this means between 12 and 14,5 per

cent of Johannesburg’s population. In mid-2008 the figure could be higher, perhaps

600 000 or 700 000. A combination of factors could be the cause of this increase:

the probable under-representation of specific types of foreigners in our survey, the

escalating crisis in Zimbabwe, and the pull factor of sustained economic growth

above five per cent in 2006 and 2007. This is not a firm finding, however, and we

should adhere to our estimate of 500 000 to 550 000 in 2006.

The foreigners in Johannesburg have significant skills to offer, although these are

more often than not experience-based rather than certificated skills.


Executive summary

Foreigners are more

than twice as likely

to be self-employed

and self-sufficient as

local adult residents

On average, the foreigners are not highly educated, although the average level of

education (incomplete high school) is often of a higher standard than equivalent

South African schooling.

The foreigners are more than twice as likely to be self-employed and self-sufficient

as local adult residents. Their level of unemployment, which occurs mainly among

very recent immigrants, is also significantly lower than the local South African

equivalent level.

The immigrants employ almost half their total numbers (12 per cent of them are

employers, employing an average of almost four people each); close to half again

are South African employees.

Up to one third of the immigrants take paid employment that unemployed South

Africans say they would like to have, but the immigrants are grateful for the oppor-

tunity, while the jobs concerned are regarded as inferior or underpaid by local

standards. In practice, there is much less direct job competition between foreign-

ers and locals than the estimate of up to one-third would suggest.

Most immigrants integrate well into the South African community, with many

South African friends, and nearly 60 per cent feel accepted as part of the local com-

munity. Most of them are happy with life in Johannesburg, rather more so than

local people.

Xenophobia is pervasive in a broad sense and is related to an accumulation of

many different negative attitudes rather than a consolidated mindset. It is caused

less by competition for jobs and resources than we expected, and at least as much

by personality attributes and ethnic loyalties among local people. However, it is

not the result of colour-based racism on the part of black or white South Africans.

The accumulated specific hostilities that make up the overall picture of xenopho-

bia do cause unhappiness among foreigners, in particular the reactions of officials

and police.

The key potential effects of hostility are muted, however, by the fact that a majority

of the foreigners interviewed feel accepted in the local community.

The national origins of immigrants are an important intervening variable, how-

ever, and people from more distant countries in Africa, excluding English-speaking

East Africa, are exposed to considerable hostility and stereotyping. The reasons are

mostly cultural, although they are sometimes expressed as an assumed proclivity

for criminal behaviour.

The hostility to foreigners does not extend to a reluctance to employ or trade with

foreigners. In fact their usefulness as employees and business people is generously

recognised by South Africans, people in business, and many key officials.

11July 2008 |

Executive summary

The hostility to

foreigners does not

extend to a reluctance

to employ or trade

with foreigners

ConclusionIn undertaking this research, CDE took on a deceptively simple but extremely dif-

ficult task of penetrating the myths and misinformation surrounding immigrant

numbers in South Africa. It is essential that policy makers have access to baseline

figures on immigrant numbers which are grounded in field research, rather than

relying on extrapolations or politically malleable guesses.

In trying to come up with such numbers CDE was all too aware that when a popu-

lation does not want to be counted (and has good reasons for this preference) then

indirect methods have to be employed, with all the potential for error that this

implies. It was against this background that we came up with a baseline estimated

figure of more than 500 000, which is substantially greater than the figure recorded

in the 2001 census, but considerably smaller than popular estimates. This would

give a percentage of foreigners in the city’s population of 12–14 per cent, depend-

ing on the accuracy of official estimates of the total Johannesburg population.

We have expressed several reservations about the findings in our analysis of the

data and review of the methodology, the most important one being that this figure

could be an underestimate.

What should we make of this estimate? It is well below the figures that have gener-

ated a kind of popular panic combined with official resignation and passivity. It

was this toxic chain reaction that contributed to the attacks on foreigners in May

this year.

However, CDE’s estimate places immigrant numbers well above any level that

could be an excuse for complacency, and our analysis clearly points to the need

for a comprehensive and effective strategy to deal with immigration issues. This

needs to be coupled with clear and bold leadership to build public confidence in

the government and the city’s ability to deal with the realities of migration.

What are those realities? As the country – particularly Johannesburg – counts the

cost of the May 2008 violence, we see three realities that will shape the future chal-

lenges of immigration policy.

The first is that continuing and probably increased immigration is inevitable. South

Africa – Africa’s richest economy – has long and porous borders at the bottom of a

poor and conflict-prone continent. Increased migration pressures, including cli-

mate change, improved transportation and political conflict will push out-migra-

tion from sending countries while the performance of South Africa’s economy will

act as a magnet for those with the initiative to leave their own countries and try for

a better life elsewhere.

The second is that at all levels, from the street trader to the boardroom, most immi-

grants make a contribution to South Africa’s economy, sometimes against daunt-

ing odds. We need to understand and manage that contribution much better, and

capitalise on it. But we will only be able to do this if we can manage flows of people

across our borders more rationally and efficiently, that is, in ways that give South


Executive summary

Popular fears and

misconceptions about

immigration must

be taken seriously

Africans confidence that these flows of people make a positive contribution to the

common good and that the country is controlling its own borders.

Thirdly, this means that popular fears and misconceptions about immigration

must be taken seriously. All over the world migration issues are disputed between

elite concerns for skilled labour needs and human rights on the one hand, and on

the other, popular concerns about preferential treatment over locals, unfair com-

petition in the labour market, border control, and bogus claims to asylum. When

they show themselves here, these concerns are not the products of blind prejudice

or some national predisposition to xenophobia, but local versions of universal and

understandable fears. None of them is as simple as it looks and all may be fanned

by misinformation. But if immigration is to make the contribution it can to our

economic growth, then they have to be taken seriously.

Neither this report, nor the research on which it is based, was intended to pro-

duce a strategy for addressing these realities. This aim belongs to a larger and more

ambitious CDE study, which covers all aspects of immigration policy but with a

particular focus on skills needs and the economy. The forthcoming final report on

this project will provide a comprehensive overview of the migration policy issues

facing South Africa, and will include research-based recommendations for more

effective policy making in the country’s approach to managing migration.

Main report

‘It is time to accept that migrants have been the

lifeblood of the city since it was founded.’

Mayor of Johannesburg, Amos Masondo, 2007

15July 2008 |

Low-level and localised

outbreaks of violence

against foreigners

have been endemic to

urban areas in South

Africa for a decade

Introduction

THIS REPORT PRESENTS the key findings of a CDE research survey of cross-border

immigrants living in Johannesburg, undertaken in 2006 and analysed in 2007.

It is being released in the aftermath of a wave of murderous attacks on foreigners in Johan-

nesburg and elsewhere in South Africa that has brought migration issues into shocking

focus, and moved migration management issues far higher up the policy agenda.

The basic facts about the attacks are well known. The violence broke out on 10-11 May in

Alexandra township in northern Johannesburg, an area of high unemployment and social

deprivation. It spread quickly to other Johannesburg townships, and then to Cape Town,

Durban, and Port Elizabeth.

By Sunday 1 June 62 people had been killed, 670 injured, and thousands displaced from

their homes. The violence subsided thereafter, but there is every reason to fear further

outbreaks.

This apprehension is based on the fact that (comparatively) low-level and localised out-

breaks of violence against foreigners have been endemic to urban areas in South Africa

for a decade. One source lists 11 such attacks on foreign nationals in the seven months

prior to the May attacks. Five of the incidents led to deaths (11 in total).1 The risk of further

outbreaks is intensified by worsening economic pressures on poor communities in urban

informal settlements, where the May outbreaks largely took place.

Aside from human and material costs in death, injury, displacement, and destruction

to property, there has been political fall-out from negative global reaction to images of

mob violence. This has fed into already worsening perceptions of South Africa which are

attributable to – among numerous other factors – political and economic uncertainty over

leadership in the ruling party, and threats to growth from energy shortages.

The government responded – too slowly for some critics – with emergency relief and

security measures, but the reactions of leaders and spokesmen as to the causes and sig-

nificance of the outbreaks were confused and sometimes contradictory. In the aftermath,

some consistency has emerged from ministerial statements. These have to some extent

acknowledged policy shortcomings and made promises of better diagnosis and prescrip-

tions for improving aspects of policy and performance in migration management.

Regrettably, none of this is new. Strains and tensions have surrounded immigration pol-

icy for more than a decade. Law-making in this area was unduly prolonged and lacked

transparency in a process that at crucial points was unpredictable and even capricious.

The Department of Home Affairs (DHA), which has primary responsibility for managing

immigration, has been plagued by capacity problems and is in a semi-permanent state of

‘turnaround’.

As a result, lack of confidence in both policy and implementation stretches from the street

to the boardroom, encompassing stakeholders and experts alike.

In the aftermath of the attacks there have been calls for change across a range of govern-

ment responsibilities, from foreign policy to service delivery. Many of the criticisms and

recommendations have been well founded. Whatever else needs to be done, however, it


Immigrants in Johannesburg

South Africa is a

typical ‘destination

country’ for cross-

border migration

will be essential to focus on the creation of public confidence in immigration policy and

the people responsible for implementing it.

To achieve this, clear, reasonable, and realistic goals for migration management will have

to be developed, as well as, feasible policies to achieve them. An obstacle here has been

the absence of credible information on many aspects of cross-border movement of peo-

ple – from the number of foreigners in South Africa to the number of permits issued by the

DHA to legal immigrants.

This dearth of well-grounded information, and the resulting prevalence of unsupported

and highly exaggerated misinformation, has probably contributed greatly to an atmos-

phere of anxiety and tension about the presence of foreigners in South Africa. Without

this, the May xenophobic attacks might not have happened.

As the attacks escalated, CDE was preparing for publication this report of a survey of local

residents and immigrants in Johannesburg. The main purpose of the survey was to try to

estimate the foreign-born population of Johannesburg through innovative methodology,

and therefore on a sounder basis than has hitherto been the case. Other aims were to

investigate the attitudes of local residents, businesspeople and state officials to foreigners,

and to profile the immigrants themselves.

South Africa as a destination

South Africa, despite being a developing country, is in some respects a typical ‘destination

country’ for cross-border migration. The differences between South Africa and its regional

neighbours, in economic opportunity, service provision, and even social welfare, resem-

ble the powerful factors of attraction that draw migrants in large numbers to Organisation

for Economic Cooperation and Development (OECD) countries from poorer neighbours,

or even further afield.

The distances over which these differentials are felt have stretched rapidly and extensively

as globalisation has advanced. While neighbouring Mexico is still the largest source of

migrants to the USA, the whole of Latin America is its neighbourhood. The EU must come

to terms with the fact that the countries of West Africa – which include some of the poor-

est, and those with the highest birth rates in the world – are in terms of migration its close

neighbours. Neither the Atlantic nor the Sahara guarantee firm borders.

The migration of people that results is difficult to manage. Among the factors which put

pressure on policy making and implementation in all destination countries, including

South Africa, are the following:

It is very hard to estimate the numbers of ‘irregular’ or ‘illegal’ or ‘undocumented’ immi-

grants (See box: Illegal, undocumented or irregular?, facing page) even in developed

societies with sophisticated resources of administration and extensive social research

capacities.

Whatever their status – legal or irregular – migrants differ widely in their motives for •

leaving, how long they stay, what they bring to the new country and what they hope to

achieve. All of these might change in the course of their sojourn.

Public attitudes to migration in the destination countries are invariably divided; •

humanitarian concerns and acceptance of diversity and skilled additions to the work-

force clash with suspicion and resentment at what is perceived to be unfair economic

competition and fears for the cultural composition of nations.

17July 2008 |

Estimating numbers and assessing impacts

Immigration policy

must perform

difficult balancing

acts between factors

that interact in

complex ways

Official responses to public divisions of this sort are often uncertainly poised between •

tougher control measures and a more flexible recognition that migration can’t be

stopped and must be accommodated in some way. Both approaches carry costs.

Illegal, undocumented or irregular?

The use of different labels for what are essentially the same people generally reflects differing

attitudes about what sort of challenges are posed by the movements of such people, and how

they should be dealt with. ‘Illegal’ has connotations of criminalisation and hopes of controlling

migration through strict law enforcement measures. ‘Undocumented’ generally goes with

conceiving migration issues in human rights terms, and managing migration through granting

status to what others call ‘illegals’ through amnesties, temporary permits, and so on. Since this

report does not make comprehensive policy recommendations one way or the other, we would

wish to avoid either extreme, at least for the moment. In the interests of consistency and neutrality,

we have opted for the label ‘irregular’ throughout the document.

CDE 2008

Key questionsIt is important to place South Africa’s concerns about migration and immigration policy in

this wider context. It is important to know that immigration policy must perform difficult

balancing acts between factors that interact in complex ways; humanitarian concerns and

international obligations; the need to control borders; citizenship and social integration;

and skills needs and the economy.

The sound policy and effective administration that are needed to perform these balancing

acts are impossible without facts and information that are as authoritative as we can make

them. In South Africa we have serious gaps in sound information in the most critical areas

of immigration, and we will find it hard to plot a strategic course through the highly com-

plex challenges of immigration management that are already upon us and set to become

even more pressing.

The list of things about which we need to know more is formidable. Among them are:

How many immigrants are in the country? We suspect that the numbers are large •

and have increased dramatically in the wake of the economic and political crisis in

neighbouring Zimbabwe, but no-one seems willing to take responsibility for the most

widely quoted figures, which are apparently without solid foundation. The govern-

ment is steadily attempting to improve its socioeconomic information on the local

population, but so far seems at a loss as to how to estimate the numbers of unrecorded

immigrants. As we shall see, this is a familiar problem in other societies as well. Do

methods in fact exist to fill this crucial gap in our information?

What is the impact of cross-border migration on our economy? What skills and experi-•

ence do immigrants bring into the country? Are they a drain on resources or do they

become economically self-sufficient? Are they temporary sojourners or do they intend

to build a future in South Africa?

Will the diversity that immigrants bring lead to social tensions and clash of cultures or •

is it enriching, strengthening historically eroded ties with our wider continent?



South Africa has

to confront closely

interlinked policy

challenges, including

a skills shortage

that is aggravated

by a brain drain

How well-deserved is South Africa’s reputation for being hostile to foreign immi-•

grants? Do sporadic outbreaks of murderous violence against particular immigrant

groups confirm this? Or is there evidence from the street and the workplace that indi-

cates a more complex and possibly even hopeful reality? What role does harassment

by bureaucrats, local officials and law enforcement agents play by comparison with

the reception foreigners receive from the general population?

Progress towards answering these and other questions is essential in view of the bigger

policy questions that challenge South Africa.

These answers could help our authorities to decide how urgent the issue of immigration

really is. The high profile accorded to unrecorded immigration may be because immi-

gration is a convenient scapegoat for domestic problems we have not yet solved – high

levels of crime, joblessness, housing backlogs and service constraints all lend themselves

to scapegoat interpretations.

More seriously, however, South Africa has to confront closely interlinked policy chal-

lenges, including a skills shortage that is aggravated by a brain drain. Do concerns over

irregular immigration and confusions about refugees and asylum discredit and obstruct

implementation of sound policies for skilled immigration? How can sound immigration

policy become part of a winning mix of strategies?

The study reported on here (see box: CDE’s research on immigrants in Johannesburg,

facing page) is one of the largest pieces of empirical research on the issue of immigration

yet undertaken in South Africa and is part of a wider ongoing investigation into immigra-

tion policy by CDE.

CDE has pursued this topic because immigration issues intersect with a number of other

crucial public policy challenges facing the country. The management of cross-border

migration is obviously highly relevant to how we deal with the issue of skills constraints

on growth. There are also policy implications for other growth-related issues like labour

market dynamics, small-scale entrepreneurial activity, and provision of social services.

Underlying all these are questions of security and public confidence in the primary gov-

ernment responsibility for controlling borders.

It is important to recognise that immigration issues affect the whole country – our Wit-

bank survey underlined this fact – and Johannesburg is not South Africa. However, the

city is the country’s most important centre of economic activity and is in some respects

enough of a defining place to warrant special investigation.

Objectives of the studyThe key focus of this research was an attempt to estimate the size of the foreign population

of Johannesburg and to expand and refine our knowledge of the socioeconomic char-

acteristics and roles of this population. These are important but ambitious and complex

aims, requiring innovative survey methodology. CDE’s methodological approach was

piloted in a smaller study in the town of Witbank2 in 2005. (see box: CDE’s research on

immigrants in Johannesburg, facing page). The methodology is described more fully in

an appendix to the main report of the survey3 on which this abridged version is based.

19July 2008 |


This report offers a

broader view of the

numbers and impact

of immigrants and

attitudes towards them

It is important to note that CDE’s survey was constructed to be broadly representative of

all areas and residence types in Johannesburg. The attacks on foreigners that took place in

May 2008 were largely confined to informal settlements bordering on major urban areas,

which represent only one area of immigrant presence. As a result, public consciousness of

immigration issues – and such research as has been done since the violence – has tended

to focus on these areas. This report offers a broader view of the numbers and impact of

immigrants and attitudes towards them. Virtually all previous empirical research among

immigrants has consisted of limited investigations of pre-selected and visible groups of

foreign residents or foreign entrepreneurs in particular locations. These studies have pro-

duced valuable findings but have had limited general applicability.

CDE’s research on immigrants in Johannesburg

CDE’s survey was developed in conjunction with an advisory team of leading South African

demographers4 for the very difficult task of reaching a population that wishes to avoid being

identified and enumerated. The approach was tested in a pilot survey of the medium-sized

industrial town of Witbank in Mpumalanga, and modified for the larger metropolitan area of

Johannesburg. The Witbank findings were reported in Immigrants in South Africa: perceptions

and reality in Witbank, CDE Focus Number 9 (May 2006).The core element of the Johannesburg

study was a household survey based on an area-stratified systematic (interval) sample, with

random starting points for intervals.

The sample was drawn within strata corresponding to the administrative districts of the

metropolitan authority, ensuring a broadly representative spread across all areas and types of

residence. It was not drawn on a racial basis but the proportions according to race emerged from

the stratified coverage of areas in which different racial and ethnic groups live. Single households

were selected on each residential stand, where necessary by sampling randomly from multiple

households on the same stand.

On the basis of this sample, one respondent was interviewed (face-to-face) in each household

in the language of his/her choice. This yielded 1002 interviews. Structured and open-ended

questions were used, the latter to allow spontaneous replies. Both direct and indirect means were

used to estimate the number of foreigners in the core sample itself, and respondents were asked

about the numbers of foreigners in the dwellings immediately surrounding theirs. An attempt was

then made to correct any invalid or exaggerated claims through a detailed probe for reasons.

Finally the responses were extrapolated to cautiously estimate the overall numbers of foreigners,

legal or otherwise.

Interviews were then conducted among foreigners whose locations were identified in the course

of household interviews, and hence were very largely based on starting points linked to the

random distribution of interviewing points in the core sample. This yielded 302 interviews with

foreign-born residents of Johannesburg.

The final element in the survey consisted of interviews with 45 representatives of business

organisations and 32 local officials. All of the officials held very senior posts in the national and

metropolitan police services, and planning divisions of local government, regional immigration

officials, and welfare.

CDE 2008



Johannesburg is

the centre of what

is by far the largest

urban-industrial

complex in Africa

The objectives of this wider and more comprehensive study of migrants in the city as a

whole were to investigate in more detail:

The approximate size, scope and patterns of origin of formal and unrecorded immi-•

gration in Johannesburg;

The validity of alternative methods of estimating broad numbers using indirect meas-•

ures designed to circumvent understandable resistance to disclosing one’s origins;

The socioeconomic characteristics of the immigrant population: occupations, entre-•

preneurial activity, levels of economic self-sufficiency;

The extent to which immigrants contribute to or subtract value from the economy and •

human resources;

The residential intentions of immigrants and their likely permanency in South Africa; •

and

The scope, nature and causes of xenophobia and other reasons for hostility towards •

foreigners.

It is important to emphasise that many of the results are based on indirect methods of

enumeration, in which representative residents’ perceptions of foreigners and identifica-

tion of foreigners in the surrounding households have contributed greatly to the body of

insights. Obviously caution has to be exercised in interpreting these results; the study was

conducted in this spirit.

Johannesburg: the cross-border magnetIn April 2007 the Johannesburg municipality opened a Foreign Nationals Helpdesk aimed

at coordinating the city’s support for migrants, providing advice on how to access appro-

priate government services, information on economic opportunities, and support for

refugees and asylum seekers. Launching the helpdesk, Mayor Amos Masondo said that

‘it is time to accept that migrants have been the lifeblood of the city since it was founded.’

The city’s policy innovation comes as escalating cross-border migration is fuelling pub-

lic concern and policy debate about the movement of people into South Africa, irregular

immigrants in particular.

The increase in numbers is being driven by the attraction of South Africa’s relatively large

economy, high per capita income, and steady growth. The push factors of political insta-

bility and precipitate economic decline are also influential, particularly in the case of

Zimbabwe, but also in more distant parts of the region, including the Democratic Repub-

lic of Congo.

These migration pressures and the highly complex challenges of immigration manage-

ment they bring, already confront policy makers in many destination countries and are

set to intensify. Johannesburg, South Africa’s economic heartland, is one focal point.

Johannesburg is a historic mining-industrial complex that has grown and diversified in

recent decades to become South Africa’s – and indeed Africa’s – Chicago, New York and

Silicon Valley, all rolled into one. As Table 1 makes clear, Johannesburg does not dwarf the

other metropolitan areas in population or disposal of public resources. However it is, in

socioeconomic terms, a ‘first among equals’. Johannesburg has a twofold significance as

a migration magnet: it is the economic centre of Gauteng province, which contributes 33

per cent to the national GDP, and, more importantly, it is in a geographic continuum of

21July 2008 |

Johannesburg’s

proximity to

neighbouring

countries is an

added attraction for

prospective migrants


close economic and social interaction with two other significant metropolitan complexes,

namely the capital, Pretoria, and the powerful industrial complex of Ekurhuleni.

In other words, Johannesburg is the centre of what is by far the largest urban-industrial

complex in Africa.

Table 1: South Africa’s metropolitan areas: population and operating budgets

Metropolitan area Population in 2007 Metropolitan government operating budget 2007/8

Johannesburg Metro 3,9 million R18,3 billion

Durban/eThekwini Metro 3,5 million R10,6 billion

Cape Town Metro 3,5 million R16,6 billion

Ekurhuleni (East Rand) 2,7 million R10,6 billion

Pretoria (Tshwane) 2,3 million R9,4 billion

CDE calculations from Statistics SA Community Survey 20075 and National Treasury data,6 2008

Johannesburg’s proximity to neighbouring countries is an added attraction for prospec-

tive migrants. It is situated 350 kilometres from the border with Botswana, 550 kilometres

from the border with Zimbabwe, and 485 kilometres from the border with Mozambique.

This is analogous to the migration routes from Mexico to the south-western United States,

and is considerably less arduous than the sea-borne hardships that African migrants

brave in their thousands to reach the borders of the EU.

For all these reasons, it is essential to look to Johannesburg for answers to the most press-

ing questions of migration management that face South African policy makers.

Towards an estimate of the number of foreigners in Johannesburg

Problems of enumerationThe starting point of all policy discussion on migration issues in South Africa has to be

the fact that there is a dearth of reliable, empirically based estimates of the numbers of

foreigners in our midst, and accurate information on their social characteristics and roles

in the economy. Estimates of the foreign-born population in South Africa have been pro-

duced – in increasing numbers as concerns have mounted over the influx of political and

economic refugees from Zimbabwe in recent months – but none of them have inspired

much confidence in their authority. This is scarcely surprising given the difficulties of

enumerating the foreign-born components of any population. Among these are:

the desire of ‘irregular’foreigners to escape detection in official surveys (censuses) or •

indeed any other kind of research sample;

the difficulty of distinguishing between foreigners who are resident (with or without •

legal status) and those whose sojourn is in some way ‘circular’ between home and host

country, involving multiple entries and exits. For instance, in 2006, there were over six



The challenges of

enumerating migrants

are even more

formidable in a country

like South Africa

million legal arrivals from the SADC countries most likely (in the popular mind) to

supply undocumented immigrants to South Africa (see box: Recorded cross-border

traffic to South Africa from African countries, facing page). This is likely to contrib-

ute to popular anxieties about ‘floods’ of immigrants and to augment (though often

only temporarily) the irregular population who stay after their visas have expired;

the ability of undocumented immigrants to acquire documents in countries – like •

South Africa – where large-scale corruption among officials creates a potentially large

class of ‘legal illegals’ which distorts numbers;

the problem of illicit documentation is exacerbated by the relative ease with which •

many migrants from neighbouring countries can assimilate in terms of language and

other markers of identity; and

large-scale population research is extremely expensive and complex to manage.•

For these and other reasons, simple head-count research, particularly on a scale big enough

to have authority and confidence about numbers, can generally be ruled out – except in

national censuses – and a variety of indirect methods of calculating numbers of foreigners

is resorted to at other times. In South Africa the combination has yielded highly fluctuating

estimates; these have usually been on the basis of official censuses or less comprehensive

official estimates that are probably least valid in the case of irregular and non-recorded

information. These official estimates of foreigners are generally regarded as questionable,

mainly because they are usually too low to match the anxieties on the ground.

The reason for the absence of credible facts on foreigners is no mystery; unrecorded for-

eigners will avoid being counted in censuses. Furthermore, there are no universally agreed

methods of enquiry into unrecorded immigration that are academically recognised, and

research of any type in the veiled world of non-legal activity is very expensive.

This lack of systematic information is not unique to South Africa by any means. Confu-

sion and political controversy reign over immigrant numbers even in the well-resourced

and well-researched United Kingdom and USA (See box: How do other countries count

irregular immigrants?, page 25).

A recent review of the challenges faced by the Office of National Statistics (ONS) in the

UK singled out migration as its main work priority and noted that, ‘It has long been recog-

nised that international migration is one of the most difficult components of population

change to measure accurately.’7 Even so, the review points out that migration statistics

are vital for policy because they are essential for population estimates and projections for

the future that are ‘used for planning, resource allocation, business decisions and a wide

range of public policy purposes’.

The challenges of enumerating migrants are even more formidable in a country like South

Africa; not only because of more limited resources and capacities than those in the devel-

oped world, but because the country faces more extensive socioeconomic problems, and

is both a sending and receiving country for international migration.

In this situation people reach their own conclusions on the basis of impressions and selec-

tive interaction with foreigners – often knowing nothing about their legal status – freely

projecting their own anxieties and interests to develop myths and stereotypes about what

foreigners mean for South Africa. In this they are often abetted by sensationalist media.

23July 2008 |

One estimate of

9,84 million irregular

immigrants in South

Africa received wide

publicity in late 2006


The result is that overly conservative official assessments of the phenomenon of non-

official immigrant numbers alternate with alarmist estimates, greatly to the detriment of

informed policy debate.

One million? Five million? Ten million?For example, one estimate of 9,84 million irregular immigrants in South Africa received

wide publicity in late 2006: ‘South Africa is home to as many as 10m (sic) irregular immi-

grants and must brace for a flood of new arrivals as its booming economy leaves poorer

neighbouring nations behind.’12 The report cited a study commissioned by a trade union

to an economics consultancy which relied largely on deportation data for its estimate.13

The figure of 9,84 million irregular immigrants amounts to 21 per cent of South Africa’s

estimated population of 47 million. Alternatively, depending on their degree of invisibility

to official survey enumerators, South Africa’s real population could be as high as 56,84

million if there were indeed 9,84 million irregular immigrants here.

In fact, in the report the 9,84 million figure is one of a range based on possible percentages

of undocumented immigrants who are deported. In this case the hypothetical percentage

is 20 per cent of the total of undocumented migrants deported; that is, the 9,84 million

figure depends on there being four irregular immigrants who evade deportation for every

one sent back. This is rated no higher than ‘possible’ according to the report’s authors, but

the figures still made headlines. In addition, no allowance appears to be made either for

circular or short-stay migration, or for the possibility that many of the deportees are multi-

ple ‘offenders’ and that there is a well-recognised ‘revolving door’ phenomenon at work.

Recorded cross-border traffic to South Africa from African countries

In 2006, 5.4 million people legally entered South Africa from the rest of Africa for trade, work, or

tourism.8 Of these, 97% came from South Africa’s immediate neighbours in SADC.9

This fact might be misinterpreted as evidence that South Africa is being ‘flooded’ with immigrants.

In fact, what this number reflects is that in many cases the same people will be making multiple

trading or work trips across the South African border each year. This can be shown by looking in

more detail at the statistics for Botswana and Lesotho. In 2006, there were 765 705 arrivals from

Botswana. The total population of Botswana is around 1.8 million people.10 This either means

that 42% of the population of Botswana visited South Africa in a single year – which is near

impossible – or that a great many people cross the South Africa-Botswana border legally multiple

times in a year. The point is even stronger for Lesotho, which has a population of 2 million and

sent 1.9 million people to South Africa in 2006.11 Either nearly everyone in Lesotho visited South

Africa in 2006 or – as is in fact the case – there are many people who cross this border monthly,

weekly and even every day.

Of course, within this pattern of circular migration, some people will choose to stay permanently,

and others will temporarily overstay their visas. But it is overwhelmingly likely that the vast

majority of the African people who enter South Africa legally leave again after a brief and totally

legitimate visit.

CDE 2008



Public susceptibility

to alarmist estimates

of undocumented

immigration is

heightened by reports

of action taken against

corrupt officials

Not surprisingly, estimated figures for migration from Zimbabwe have received public-

ity in South Africa. For example, in a recent report, the Zimbabwe Central Bank claimed

that 1,2 million Zimbabweans have migrated to South Africa since 1990.14 How many have

returned or been deported is not estimated, but uncorroborated estimates by some South

African government spokesmen that 3 million Zimbabweans are currently living in South

Africa have become accepted public wisdom. A figure of 3–4 million Zimbabweans in the

diaspora has received wide currency, without appearing to have much in the way of sta-

tistical authority behind it. One NGO report quotes a Zimbabwe government spokesman

as saying that ‘60-70 per cent of Zimbabwean adults who should constitute the productive

population are living abroad.’ The figure cited this in support of an estimate of 3,4 million

Zimbabweans outside the country, with the assumption that most are in South Africa.15

As we shall see later in this report, there are grounds for questioning whether such a

large presence (especially of a single nationality) is compatible with what we know about

the demography of South Africa. There are also grounds for scepticism based on popu-

lation figures for Zimbabwe. As Loren Landau points out,16 about 3,6 million of Zimba-

bwe’s population of 12,3 million are adult males, who are the most likely migrants: ‘As it

is unreasonable to assume that all adult Zimbabweans have left the country, there is an

upper cap on the number who might have come to South Africa.’ Since the pool of movers

is shared between Zimbabwe’s other regional neighbours and more distant destinations

like the United Kingdom and United States, serious scepticism about the figure of 3 mil-

lion in South Africa is called for.

Certainly it is clear that deportations from South Africa are rising. The 2003–2004 annual

report of the DHA put deportation statistics for all nationalities at 164 294. In 2004–2005

the figure had only increased to 167 137, but by 2005–2006 it had jumped to 209 988, a rate

it sustained into 2006–2007, when the total was 266 067.17 It is unclear what these figures

mean; the trend might partly be due to increased capacity to detain and deport unre-

corded immigrants and to the ‘revolving door’ phenomenon. In any event, the swelling

numbers suggest that the volume of unrecorded immigrants is increasing substantially.

Public susceptibility to alarmist estimates of undocumented immigration is heightened

by reports of action taken against corrupt officials. It is widely recognised – by the DHA

among others – that this form of corruption has made it very easy for many otherwise

unrecorded immigrants to purchase the documentation required to ‘legalise’ their stay

in the country. For instance in late 2007, 15 officials at the Caledonspoort Port of Entry

in Fouriesburg, Free State province, which borders Lesotho, were arrested on corrup-

tion charges related to abetting the entry of irregular immigrants in to South Africa.18

During 2006 and 2007 the Chief Directorate Counter Corruption and Security of the DHA

concluded 125 fraud and corruption cases, although it is not clear how many of these

related to undocumented immigrants.19

In contrast to reports and accounts such as these, estimates which are more securely

grounded in population statistics usually converge on those made on the basis of the 2001

census. According to the census, around 1,1 million people declared themselves to have

been born outside South Africa, which amounted to 2,3 per cent of the country’s popula-

tion at that time. In Johannesburg, the census figure for these individuals was 216 715 –

6,2 per cent of the city’s population. This was close to three times the percentage for the

country as a whole.20

25July 2008 |

Censuses all over the

world are notorious

for undercounting

people who would

prefer to be invisible


The 2001 census must be treated with extreme caution. Censuses all over the world

are notorious for undercounting people who would prefer to be invisible for whatever

reason.

There are thus no reliable estimates of even the broad numbers of cross-border immi-

grants in South Africa at present. Almost everyone agrees, however, that the numbers are

large. Given the fact that despite its relative wealth in an African context, South Africa is

itself struggling to cope with high levels of unemployment and shortfalls in resources to

How do other countries count irregular immigrants?

Counting irregular immigrants is an extremely sensitive and involved process, where the counter

must, by nature of the subject, make estimates, presume, extrapolate and standardise. The United

States uses the residual method to calculate the number of irregular immigrants in the country.

This method subtracts the number of immigrants who are authorised to be there from the number

of foreign-born residents counted by the Census Bureau through their annual Current Population

Survey report, then adjusts the number using estimates of immigrants’ deaths and out-migration,

and for census undercounting.

This figure is supplemented by analysing trends in foreign remittances and school enrolments in

localities with high populations of irregular immigrants. Also, the figures that the residual methods

reveal are compared with population deficit numbers from countries with high immigrant sending

rates (such as Mexico).

However, the limitations of this method are worrying. The residual method doesn’t take into

account the student or other long-term (but non-immigrant) population. It also doesn’t differentiate

between irregular migrants, and those on their way to legalisation (waiting for a green card,

temporary protected status), and the United States government doesn’t offer a margin of error

for these calculations.

Across the Atlantic, the United Kingdom has considered several options for estimating its irregular

population. A variety of indirect methods could be used to measure this, all of them based on

the premise that irregular immigrants will be recorded in some way during their period of

residency, through police or employment records, birth/death/fertility data, employment records,

or by monitoring the sex/age ratio (some of these subject to comparison with the same data

from suspected sending countries). A popular choice for qualitative data is the Delphi method:

surveying a range of people, and then using the answers to obtain an average estimate.

The Netherlands has used the ‘capture-recapture’ method, based on police records of

apprehended irregular immigrants, while Switzerland sent 5 500 anonymous questionnaires to

employees of employers who were known to hire asylum seekers. Numerous problems arise from

these methods. Using different data sources to obtain a ‘bigger picture’ of irregulars in a country

is problematic, as such data sources are not collated(for instance, there is no merging of prison

population versus court appearances), or not detailed enough. The Delphi method often results in

a ‘census of opinion’ which cannot be statistically and reliably verified. The Netherlands’ focus on

criminal records has a bias towards those irregulars with a good chance of getting caught (those

with poorly forged documents, those who are engaged in criminal activity or who do not have a

strong social safety net), while Switzerland’s questionnaires presume a certain level of knowledge

from the respondent, and also presume respondents would answer truthfully and altruistically.

CDE 2008



There are thus no

reliable estimates

of even the broad

numbers of cross-

border immigrants in

South Africa at present

service the needs of its poor people, the issue of immigration flows is of vital importance

in any strategic assessment of South Africa’s socioeconomic challenges.

It is appropriate, therefore, to begin the analysis by considering the results of the enquiry

into the estimated numbers of foreigners in Johannesburg. In various ways these esti-

mates inform much of the substance of the rest of the investigation. The results will also

shed further light on the extent to which it is effective to use partially indirect results in

empirical surveys as a basis for estimating numbers of unrecorded foreigners in a city or

country.

Foreigners in the base sampleThe most obvious and certainly the least controversial way of estimating the numbers of

foreigners in the city is to make calculations based on the core sample. Hence, early in the

interview, we asked the respondents which countries (other than South Africa) they knew

well, and in which they had lived and worked. The intention was to provide initial indi-

rect indicators of foreign origins among types of people who, if they were South Africans,

would be most unlikely to have travelled outside the country. Table 1 shows the results.

Table 2: Respondents’ acquaintance with foreign countries

Proportions of respondents who knew one

or more foreign countries well or had lived

or worked in foreign countries

Black

sample Minorities

Total

sample

Knows foreign countries ‘well’ 19,4% 60,5% 28,7%

Has lived or worked in a foreign country 7,1% 28,1% 11,9%

CDE Johannesburg Survey 2008

These results must be cautiously interpreted. The large number of people in the minor-

ity sample who claim to know other countries well could be a reflection of active mid-

dle-class travel and tourism patterns, although it undoubtedly does include substantial

immigration. In the black component of the sample, where only about 7 per cent would

approximate middle-class status, the 19 per cent claiming to know other countries well is

surprisingly high. However, many of these people could have been basing their claims on

media accounts.

Living and working in a foreign country, however, comes a little closer to being an index

of foreign origin in the black sample, among which very few people are of the type to have

left South Africa to live or work elsewhere and then returned. The 7 per cent in the black

sample who claimed to have lived and worked in a foreign country is high in a sample

of predominantly poor or newly upwardly mobile people, and therefore probably does

indicate foreign status among most of them, say around 5 per cent of the sample. Most of

the countries mentioned in this context were BLNS countries, suggesting that the foreign

respondents were still very cautious early in the interview – BLNS countries are so inter-

woven with South Africa that it would have been reasonably safe to mention them.

We have no doubt that 5 per cent is an underestimate of foreign origin among blacks,

because not all foreigners who might be unregistered or dubiously registered would admit

their citizenship or immigrant status, even in this indirect way.

27July 2008 |

As many as 40 per

cent of inner-city

respondents were

judged to be foreign


Our next procedure was to ask the interviewers to record their impressions of the origins

of people in the sampled households. The experienced interviewers were all proficient in

several languages and they had conducted up to thousands of interviews in the Gauteng

area. They were therefore in a good position to provide reports on respondents whom

they considered to be foreign. Their judgements were not speculative and were based on

discussion of the respondent’s history. The results of the confidential interviewer reports

were as follows:

In the black sample the interviewers considered 12,4 per cent of the respondents to be

foreign, and among the minorities 6,6 per cent. In table 3 the results of the interviewer

observations are given for the entire sample:

Table 3: Interviewer observations and judgements about the foreigners in the total core sample

Interviewer observations or judgements Results

Proportion who admitted that they were immigrants 8,6%

Proportion in addition to those above who spoke a foreign language or spoke with a clearly foreign accent and whom the interviewers considered to be foreign 2,5%

Overall proportion of respondents who were judged to be foreign 11,1%

Breakdown of ethnic identities in the sampled foreign respondents: (% of foreigners)

Zimbabweans

Mozambicans

BLNS countries

British

Europeans

Nigerians

Malawians

Somalis

Egyptians

From elsewhere in Africa

39%

14%

15%

7%

5%

4%

4%

1%

1%

8%

CDE Johannesburg Survey 2008.

As many as 40 per cent of inner-city respondents were judged to be foreign, a proportion

that dropped to 8,1 per cent of shack dwellers and 7,5 per cent of suburbanites.

The details given are estimates of the proportion of foreigners in the drawn sample. They

are likely to be on the conservative side, because of their reluctance to be interviewed in

a regular sample. Some will claim that they are not members of the household, ‘Ignore

me, I am just a visitor here’ etc, while others will simply not respond at all, either refusing

or claiming to feel ill, etc. To allow for this, two further strategies were followed. These are

described below.

Estimates based on surrounding householdsFirstly, within each household sampled, after thorough reassurances of confidentiality,

we asked the respondents and other available well-informed members of the household

to think of the four dwellings next door (two on either side), their own dwelling, and the

five dwellings roughly opposite. We asked them to fix these dwellings in their minds. There

was an element of estimate in the identification of the 10 dwellings, due to street corners,



The issue of

immigration flows is

of vital importance

in any strategic

assessment of South

Africa’s socioeconomic

challenges

vacant houses and other peculiarities of layout, while in the poorer areas there were quite

often additional houses or shacks in the grouping of dwellings discussed. However the

dwellings that the respondents fixed in their minds constituted ten or more closely clus-

tered adjacent dwellings within sight and often sound of the respondent’s dwelling. The

exact number of dwellings being considered was recorded and interviewers provided a

sketch map of the layout of relevant dwellings.

Then we asked how many adults of 16 years and older living in the group of dwellings,

including their own, and including lodgers and people living in outhouses and garages,

they knew or considered to be foreigners or immigrants. We also asked them why they

thought so.

Finally, we asked them for the overall number of adults 16 years and older in all the dwell-

ings under consideration. We generally tried to avoid speaking about a specific dwelling

because of the understanding of confidentiality that we had established. The unit of dis-

cussion was the cluster of dwellings.

On the whole the co-operation was very good. In the sample as a whole, 98 per cent were

able and willing to identify the numbers of dwellings in the surrounding clusters, 93 per

cent were able and willing to estimate the numbers of foreigners in the dwellings, if any,

and 86 per cent were able to estimate the total number of adults in the dwellings. Despite

the high levels of response, in the informal areas there was a degree of tension relating

to this question, because of the close proximity of houses to one another. Although the

discomfort did not prevent most of them from giving estimates, it has to be noted that the

responses could have been downwardly biased by this discomfort.

The second strategy was that after asking for the number of foreigners in the surrounding

houses, the respondents were asked what particular kind of foreigners they were (nation-

alities). They then had to become more systematic and itemise foreign nationalities in the

clusters. As part of this exercise they were asked how many there were of each nationality.

The numbers of each nationality were then aggregated to arrive at an additional estimate

or check on the previous estimate of the number of foreigners.

The results are summarised in table 4. Here we should point out that in the interests of the

greatest possible precision, at a level not necessary in the other aspects of the analysis,

we weighted the sampling results in black areas to make sure that formal and informal

housing were in the correct relative proportions as indicated by the outdated 2001 Johan-

nesburg statistics.

Note that the second estimate of the mean number of foreigners in the clusters is well

above that for the first estimate – see table. There could be three reasons for this.

First, the task of itemising nationalities and then estimating numbers made the respond-

ents more systematic, and probably more inclusive than they were in the first estimate.

Second, the respondents were probably relieved that after the first estimation they were

not asked for the foreigners’ addresses. They were therefore more relaxed and trusting in

the second estimation exercise.

Thirdly and most importantly, however, not all the people could or would specify the exact

nationalities of foreigners. The rate of participation therefore fell, and with it the active

sample on which the second estimate is based. Put differently, relative to the response

rates in the first estimation exercise, far fewer respondents felt confident in identifying the

nationalities of foreigners, and hence far fewer were able to estimate a total of all types of

29July 2008 |


foreigners. The resulting changes in the sampling figures probably account for most of the

difference between the outcomes of the two sets of estimations. We thus had a problem to

reconcile the two sets of estimates in table 3. We decided to proceed as follows.

The first estimate in table 3, the total number of foreigners per cluster irrespective of

nationality, may be accepted as a statistically adequate estimate because the response

rates are rather high at roughly 94 per cent. Hence there is not likely to be a significant

response rate bias in the estimate of 8,6 foreigners per cluster that emerged.

The second estimate, which is based on a response rate of only 41,3 per cent, cannot be

accepted as is and has to be adjusted in some way. We suggest a midpoint between a mini-

mum and a maximum figure, based on certain assumptions. If one assumes that in all

cases where people could not respond there were no foreigners in the clusters at all, then

we have a theoretical minimum. On the other hand, if one assumes that in all these cases

of non-response the true number of foreigners was the same as that estimated by the peo-

ple who did respond, we have a theoretical maximum. The theoretical maximum would

be 23,68 foreigners, and the theoretical minimum would be 9,79 foreigners (23,68 x 494

Table 4: The steps and calculations used to arrive at estimates of the proportions of foreigners in Johannesburg

Type of

residential area

Base sample size

(households

weighted, see text)

Mean number

of dwellings

in surrounding

clusters

Active sample

(numbers able

to respond to

questions on

numbers of

foreigners in

clusters)

Mean number of

foreigners per

cluster

Mean total number

of adults per

cluster

First estimates

Suburbs: blacks 54 10,24 49 2,71 36,88

Suburbs: minorities 226 9,59 200 2,27 23,79

Domestics 64 10,19 29 0,66 11,91

Formal townships 384 12,32 382 1,55 38,73

Informal areas 347 15,40 345 3,70 31,39

Inner city: blacks 118 20,95 116 61,59 153,17

Inner city: minorities 2 7,00 2 19,50 172,50

OVERALL 1 195 13,29 1 123 8,60 46,30

Second estimates

Black areas 967 14,15 425 26,35 49,92

Minority areas 228 9,56 69 7,25 25,68

OVERALL 1 195 13,29

494

(missing cases 701) 23,68 46,30




Both the theoretical

minimum and the

theoretical maximum

are highly unlikely,

and the truth probably

lies somewhere

between the two

+ 701x 0 /1 195). Both the theoretical minimum and the theoretical maximum are highly

unlikely, and the truth probably lies somewhere between the two. We have no recourse

but to take the midpoint, which would be 16,74 foreigners. It is appropriate that the esti-

mate be higher than the 8,6 foreigners, because of the more systematic method used.

Further adjustments to these estimates may be necessary, however. The major problem

with the ‘identification’ of foreigners by local residents is the danger of over-estimation of

foreign numbers because many people who look unfamiliar may be regarded as foreign,

even people from distant regions within South Africa. How likely is this to have been the

case in this exercise?

After asking for estimates of the foreigners living in the clusters of surrounding houses,

we proceeded to ask the respondents for their reasons for thinking that the people identi-

fied were foreign. This question was asked of each type of foreigner identified. The overall

distribution of reasons is given in table 5, in descending order of likely validity.

Table 5: Reasons for considering the people in surrounding houses to be foreigners (multiple replies), weighted by the numbers of types of foreigners to whom the reasons apply

Weighted reasons Blacks Minorities ALL

Has been informed by foreigner 26% 17% 24%

Has been informed by landlord or by wife/partner of foreigner

14% 8% 13%

Speaks a foreign language not known in SA 21% 26% 22%

(Africans’) poor command of or inability to speak any South African languages except English

16% 3% 13%

Speaks with distinct foreign accent 9% 10% 9%

Clothing and style of dress 7% 15% 9%

Features and appearance 17% 28% 20%

Culture and behaviour 15% 35% 20%

Types of employment 5% 16% 8%

Does not mix with South Africans/hides away 1% 8% 3%


We must remember that interviews received multiple replies, and it often happened that

both more and less plausible reasons were given for identifying particular foreigners or

types of foreigners. The first four reasons are the most plausible and would be taken as

good indicators of foreign identity by most people. They may be false indicators in some

cases, but this bias would in all probability be counterbalanced by valid identifications on

the basis of less plausible reasons. The first four reasons together accumulate to some 72

per cent of the reasons and (drawing from detailed cross-tabulations) about 66 per cent of

the identifications. This would indicate that about two-thirds of the identifications were

made on a reasonably plausible basis, and about 34 per cent on a less secure basis. (Here

we must bear in mind, however, that West and North Africans themselves state that they

are physically different from local Africans.)

31July 2008 |

Zimbabweans

and people from

BLNS countries are

underestimated in

the CDE Survey


Thus the two mean estimates of the numbers of foreigners in the clusters should perhaps

be reduced by around 34 per cent each to eliminate the effects of less plausible identifica-

tions, yielding downwardly adjusted estimates of 5,68 and 11,05 respectively.

These estimates need to be related to the estimated mean numbers of adults both foreign

and local in the clusters (46,30). Respectively, they would be 12,27 per cent and 23,87 per

cent of adults in the sample.

Estimating the size of the foreign presence in Johannesburg: estimates of numbers of foreigners compared and discussedWe now have a basis for comparing all the estimates derived so far, comprised as follows:

Foreigners as an estimated proportion of adults in the base sample: 11,1%•

Foreigners identified in surrounding residential clusters: •

Overall: 12,3%

Aggregated nationalities: (midpoint estimate – see above) 23,9%

These estimates are all in the form of proportions of adults. Grossed up as estimated num-

bers of foreign adults among the 2 218 000 adults in Johannesburg (as given by Statistics

SA, Statistics in Brief, in 2005):

246 2421.

272 8632.

530 1893.

These estimates are all subject to qualifications and statistical reservations but together

they provide a basis for at least an informed assessment of the order of magnitude of the

foreign adults. But the range of variation is wide. Are the numbers likely to be closer to

the lower estimates of around 250 000 to 270 000, or to the higher estimates of 500 000 to

550 000 or more?

First we must assess the likely bias in the estimates. There is one major reason for suspect-

ing that the lower limit of 250 000 is too low, ie that where foreigners have been identified

by neighbours, they have usually been either legal immigrants of long standing or have

been part of clusters of foreigners where there was no danger of victimising individuals in

answers to our questions – therefore the more isolated foreigners dispersed through the

fabric of residential areas are most certainly underrepresented. In this context the types of

foreigners identified in the surrounding households is informative. See table 6.

It is important to note the following. The estimated numbers of Nigerians far exceed the

numbers of Zimbabweans, Mozambicans and people from neighbouring BLNS coun-

tries. This rank order of numbers for all foreigners in Johannesburg is highly unlikely. It

is precisely the Zimbabweans, people from Mozambique and those from the neighbour-

ing BLNS countries who are best able to disperse themselves throughout the fabric of

residential Johannesburg because their languages (isiNdebele, Xitsonga and Sotho and

Nguni-based languages) allow them to interact freely at grassroots level and, in the case

of Zimbabweans and those from the BLS countries, are quite soon able to pass them-

selves off as South Africans. They can work as gardeners and domestics and simply not be

noticed. There can be little doubt that Zimbabweans and people from BLNS countries are

underestimated in all the numbers produced above. Given the proximity of these coun-



Serious scepticism

about the figure

of three million

Zimbabweans in South

Africa is warranted

tries, and the socioeconomic conditions in Zimbabwe, they can be expected to represent

the largest numbers, and certainly not to trail behind Nigerians.

The Zimbabweans are also very poor in general and least able to establish themselves

in visible households of their own, or rent rooms in houses, and hence were also under-

represented in the very first estimate based on foreigners in the core sample of house-

holds. If they had been fully covered in the estimates, therefore, the numbers in the lower

estimates would certainly have been significantly higher. CDE is confident that the lower

grossed-up numerical estimates (circa 250 000) err significantly on the conservative side.

Does this mean that the higher estimate of circa 500 000 is probably roughly correct? The

major reason for suspecting it to be too high is that people identify anyone who looks and

sounds different as a foreigner. But we have already corrected for this possibility – the two

highest estimates produced by the methodology were reduced by 34 per cent to broadly

eliminate identifications based on mere appearance, culture and lifestyle. Furthermore,

the higher estimate was also reduced to allow for non-response.

On these grounds, CDE is confident that the grossed-up numerical estimate in the range

of 500 000 to 550 000 is more likely to be correct than the lower figure. In fact, if anything,

even this estimate is likely to be too low, because it clearly under-estimates Zimbabweans

and BLNS citizens relative to Nigerians (see again table 6 above). Johannesburg seems to

be home to at least half a million foreigners – and it could be a couple of hundred thou-

sand more. This suggests at minimum a foreign presence in the city more than twice that

estimated by the 2001 census and an escalation from 6,72 per cent to more than 14 per

cent.

Whatever the relative accuracy of the various estimates presented, they are all subject to

a likelihood of under-estimation. We have already briefly noted the sources of under-esti-

mation, but they are sufficiently important to be set out in detail. There are three catego-

ries of foreigners that are very probably under-represented to a lesser or greater degree in

Table 6: Number of types of foreigners identified in all the clusters surrounding the sample dwellings

Types of foreigners Blacks Minorities ALL

Nigerians 4 410 112 4 522

Zimbabweans 3 216 39 3 255

Mozambicans/Angolans 1 784 30 1 814

Malawi and Zambia 440 18 458

Somalia and rest of Africa 378 24 402

East Africa 304 36 340

Neighbouring BLNS countries 168 21 189

All people from the East 91 128 219

All European countries/USA/Australia 30 141 171

Total 10 821 549 11 370


33July 2008 |

The figure of half a

million foreigners

in Johannesburg in

2006 is probably a

conservative estimate


all our estimates of foreign numbers, and also in our results from the foreign interviews.

They are the following:

Those whose current vulnerability to arrest and deportation is so great that they 1.

choose to lead a highly reclusive life, avoiding contact with neighbours and so avoid-

ing recognition as foreigners.

Those at the opposite end of the scale whose integration into the local community has 2.

been so successful that they are no longer recognised as foreign and therefore were

not identified by neighbours in the first place. This applies particularly to immigrants

from Britain, the Commonwealth and parts of Europe, to Indians and to Southern

African immigrants whose ethnic cultures are close to those of the locals.

Those whose preferred residence is in the backyards of suburban houses and urban 3.

fringe smallholdings (usually as unregistered gardeners or agricultural labour) and

who consequently are not located in visible concentrations. This applies particularly

to Zimbabweans, Malawians and some other Southern African immigrants.

On this basis our concern should not be about over-estimation of numbers, but rather of

underestimation. In this light the figure of half a million foreigners in Johannesburg in

2006 is probably a conservative estimate by a margin that is difficult to estimate but could

be substantial.

Assessing impacts: the balance sheet as seen by residents, immigrants and officials

Besides questions related to enumerating the foreign presence in Johannesburg, the sur-

vey interviews had several other focal points:

Johannesburg residents were asked about their perceptions of where foreigners live; •

how they recognised them; their perceptions of foreigners’ contributions to the econ-

omy; their levels of acceptance or rejection of foreigners.

The immigrants were asked questions aimed at constructing a profile (gender, age, •

length of stay, country of origin, etc) and about roles in the economy and experiences

of acceptance or rejection.

Officials and representatives of business organisations were asked about their percep-•

tions of immigrant numbers and their impact on the city. Questions were also asked

about business experiences of skills shortages and immigration as an approach to

addressing skills issues.

What do Johannesburg residents think of the foreigners in their midst?How do they recognise foreigners? Where are they from? Where and how do they live? Are

they a problem or an asset?



The stress of crime

creates a perpetual

need among people

in Johannesburg to

find scapegoats

RecognitionResidents recognise foreigners on the basis of multiple indicators – on average between

2 and 3 different indicators, including controversial criteria like appearance, habits and

physique. However, three-quarters of the recognition among black people in our sample

is based on hard knowledge, and 45 per cent of the recognition among minorities has a

similar foundation. Almost nobody identifies a foreigner on the basis of assumed physical

and cultural characteristics only.

Johannesburg residents seem to agree remarkably on the relative numbers of immigrants

in their midst. Mozambicans and Zimbabweans are the clear frontrunners, and this col-

lective image is empirically plausible. Since the survey fieldwork occurred, a deepening

economic crisis in Zimbabwe has probably lifted Zimbabweans above Mozambicans.

After the two leading groups, sheer numbers seem to alternate with cultural or ethnic vis-

ibility in determining the impact on Johannesburg residents. Nigerians seem to occupy

third place in terms of impact, based on both a strong numerical presence and cultural vis-

ibility, followed by Chinese and Malawians, and then at a lower level by Indians, Pakista-

nis and Europeans. The number of people from the surrounding BLNS countries probably

exceeds that of Nigerians, but they have very low visibility and cultural distinctiveness.

Somalis have low numbers but strong visibility. Immigrants from elsewhere in Africa have

relatively low visibility, due mainly to smaller numbers and a tendency to cluster.

Johannesburg has long been a magnet for in-migration from other provinces in South

Africa, lifting its population by up to 10 per cent between 2001 and 2006.21 Despite the sig-

nificant numbers of local migrants, 60 per cent of black Johannesburg residents perceive

foreign immigration to be larger than local in-migration, a perception shared by 40 per

cent of minorities. There is clearly a substantially shared perception of massive foreign

entry.

LocationThe dominant foreign groups – Mozambicans, Zimbabweans and Nigerians as well as

significant numbers of Somalis – are perceived to cluster in the inner-city areas of Johan-

nesburg, including Hillbrow and Joubert Park. These areas are also the first places that

immigrants go to, and it seems that many of them move out once they have familiarised

themselves with the city.

As one moves to the fringes of the inner city areas, so the composition of immigrants

diversifies. Further out, the formerly white suburbs have a more selective immigrant pres-

ence, accommodating significant numbers of Nigerians, Chinese, Indians, Pakistanis, and

Europeans.

The Zimbabweans, Mozambicans and Nigerians are also well spread in the outlying areas

of the city. The numbers of Zimbabweans and Mozambicans are so large in relative terms

that they also have a strong presence in shack areas and townships. They are the most

dispersed of all immigrants.

Nigerians and other West Africans have not hesitated to seek accommodation outside the

areas and types of accommodation in which immigrants from Africa usually cluster. These

West African groups, with some North Africans, have joined the Asians to lead the way in

breaking the racial mould of the post-apartheid city.

35July 2008 |

Two thirds of officials

interviewed concede

that foreigners

are hardworking,

determined and

productive


WorkJohannesburg residents have good collective vantage points to observe the labour and

work of the immigrants. In order of relative importance, this activity is perceived to break

down as follows:

Hawkers1.

Artisans and skilled trades2.

Miners and other semi-skilled labour3.

Domestics and gardeners4.

Shopkeepers and stallholders5.

Professional, technical and semi-professional activity6.

Personal services7.

Other service occupations8.

Security guards9.

Small manufacturing and craft work.10.

The residents also think that some 14 per cent of immigrant activity is illegal or criminal,

but this could very well be a projection of their anxieties rather than well-founded empiri-

cal observation.

Acceptance/rejectionRoughly one-third of Johannesburg residents regard foreigners as generally ‘bad for the

city’. This is no worse than the image of low-income foreigners in most countries of the

world. It is also significantly lower than the level of negative perceptions found in our

previous study in the smaller town of Witbank. In this regard Johannesburg residents are

responding very much as one would expect of people in a large cosmopolitan city.

Overall rejection may be at tolerable levels, but when it comes to specific aspects of the

image of foreigners, however, the picture in the minds of local residents becomes more

negative. Perceptions of specific negative features of foreigners outweigh positive features

by a factor of 1,25:1 among black residents and 1,95:1 among minorities. These negative

perceptions are substantially centred on suspected criminal tendencies among immi-

grants, including the fact that large proportions of immigrants are believed to have ‘nor-

malised’ their residence in Johannesburg through bribing corrupt officials, as confirmed

by the publicly expressed concerns of high-ranking government officials. Other important

negative images are that foreigners compete for scarce urban land, housing and services,

and, at a somewhat lower level of frequency, their image of competing for job and small-

business opportunities.

We constructed an index of xenophobia based on a mix of evaluations of the assumed

social and moral character of immigrants, and found that whites were least hostile and

Indians most hostile to immigrants, with blacks and coloured people in between.

Other than this, xenophobia is not consistently based on the expected factors like per-

ceptions of economic threat and competition among poor local people who might feel

that their jobs and life chances are threatened. Neither can xenophobia be related to dis-

guised racism. Higher levels of education also do not consistently lead to more positive

perceptions of foreigners, except among middle-class minorities, among whom a rela-

tively benign liberal mindset leads to a more welcoming attitude to people perceived as



One pervasive factor

in xenophobia is crime

underdogs. Even the length and quality of contact have a complex relationship to hostility

– the attitudes of middle-class liberals tend to sour with more intensive contact, whereas

blacks’ attitudes improve, but the effects are generally weak.

Is xenophobia stimulated by the role and activity of foreigners in the city? Immigrant

efforts to improve their circumstances, the skills and occupational experience that they

bring to their new city, and even their contributions to job creation and the economy,

do not ensure immunity from xenophobic hostility. Many of our respondents, black and

white, were quite capable of saying that they felt that the presence of foreigners was dan-

gerous, but that they would prefer employing some kinds of foreigner to local people.

Xenophobia unpackedOne pervasive factor in xenophobia is crime – not solid evidence about the crime that for-

eigners may or may not be committing, but the fact that the stress of crime levels creates

a perpetual need among people in Johannesburg to find scapegoats. Crime, as a conse-

quence, has negative associative effects and has created a self-perpetuating myth that a

foreign presence is tantamount to a criminal presence.

Variations in the scores on our xenophobia index suggest that higher levels of hostility to

foreigners are associated with higher incomes and managerial positions among blacks,

but lower level incomes and occupations among minorities. Among minorities, mem-

bers of the Pentecostal churches and Afrikaans-speakers generally have somewhat higher

levels of dislike of foreigners. Among blacks it is the emerging middle classes and the

English speakers who have the greatest dislike of foreigners. Other major factors include

personality attributes of mistrust and misanthropy among minorities, and feelings of rela-

tive deprivation among both blacks and minorities. A Factor Analysis shows that, more

often than not, being anti-foreign is not a consolidated mindset, but an accumulation of

different kinds of hostilities to varied categories of foreigners, and that material threats

and competition are often less important than ethnic affinities and psychological vari-

ables in shaping anti-foreign sentiment. Sometimes anti-foreign hostility is not related

to anything in particular – except conformity to views expected of people and a shallow

acceptance of urban legend.

Who is accepted in Joburg?Despite all the generalised hostility, there is surprising agreement among Johannesburg

residents, both blacks and minorities, that certain categories of immigrants are beneficial

for Johannesburg. In three broad categories of approval the immigrants are:

Highly approved: BLNS citizens, Westerners, Indians•

Approved: Chinese, Zimbabweans, Mozambicans•

Approved with qualifications: Malawians, Pakistanis•

These people have a reputation for contributing to the economy, for being highly pro-

ductive and willing workers, and for having noteworthy skills and education. While not

cancelling certain xenophobic reactions and mixed feelings of various kinds, most Johan-

nesburg residents will concede that there is a place for all these immigrants in the city.

Nigerians, Somalis and people from non-English-speaking African countries (except

37July 2008 |

Two-thirds of

the people in the

sample come from

either Zimbabwe or

Mozambique, with the

rest evenly spread in

terms of their origins


neighbouring Mozambicans who belong to ethnic groups found in South Africa) do not

achieve this status of approval.

Immigrants speak for themselves: a profile based on immigrant interviews

Unfortunately, the results of the follow-up survey among 302 foreigners cannot give

a complete picture of the foreign residents of Johannesburg because, as already noted,

the results underestimate the foreigners who live and work on suburban and peri-urban

properties, as well as former residents of the BLNS countries and immigrants who live

in non-residential circumstances. Nevertheless, the results are a good representation of

those in the residential fabric and who are less reclusive.

One general observation on survey feedback is crucial and should be made at the out-

set. In a large number of the interviews with foreigners it became subtly apparent that

the respondents were neither ‘illegal’ nor fully legalised in terms of formal entry require-

ments. Details were obviously never stated, but it seems that widespread corruption

among petty officialdom has created a new category of people who will probably go on

to become proudly South African – people who have purchased their rights of residence,

or perhaps even South African identity documents from corrupt officials. Although it is

not possible to quantify this suspicion, it had a sufficiently strong and frequent effect on

experienced interviewers to be worth noting.

The foreigners in our sample are concentrated in informal housing areas (41 per cent)

followed by a variety of types of accommodation in formal townships (23 per cent) with

the remainder more or less evenly divided between the inner city and the suburbs. Had

we included more recently arrived foreigners, the inner city would have gained, because

its anonymity makes it an important receiving area for new arrivals.

We cannot repeat the full profile presented in the body of the report, but a few significant

features of the foreigner profile stand out:

They are predominantly male (64 per cent) and 25–49 years of age.•

The median length of stay in Johannesburg is 5 years, but no fewer than 14–15 per •

cent have been resident for less than two years. The tempo of arrivals is generally not

decreasing and it should be assumed that it is increasing sharply at the present time,

with accelerated entry from crisis-ridden Zimbabwe.

Two-thirds of the people in the sample come from either Zimbabwe or Mozambique, •

with the rest evenly spread in terms of their origins.

When it comes to home visits, the sample is largely divided between those who visit •

their homes one or twice a year, and those who never leave Johannesburg.

About half live in shacks, either in informal areas or in townships, and the rest are •

spread across other types of accommodation.

They have a wide spread of educational qualifications, but the largest single group has •

an incomplete high school education.

Of the African immigrants 75 per cent earn less than R2 000 a month, but the people •

from the West and Asia have middle-class incomes. Hence the South African race–

class division is repeated among foreigners.

Some 80 per cent are in permanent or semi-permanent relationships, mostly with other •

foreigners (59 per cent), and more than 60 per cent have children in South Africa.



Only some 20 per

cent of the foreigners

interviewed are

unemployed and

looking for work, and

they have been here

the shortest time

The majority speak English and major African languages either fluently or very well.•

Work: immigrants in the economy

Only some 20 per cent of the foreigners interviewed are unemployed and looking for

work, and they have been here the shortest time. A high proportion of some 44 per cent

of the immigrants are self-employed, relatively far more than the 12–16 per cent of South

African adults in Gauteng who are self-employed. Some 12 per cent of the immigrants

employ people, an average of nearly four staff members in their operations, almost half

of them South Africans. This means that the number of people employed by immigrants

would amount to almost half the immigrant numbers.

The list of occupations yielded by interviews of the sample of foreigners was much the

same as the list of perceived foreigner occupations in the core household sample. There is

clearly a great deal of experience-based skill among the foreigners. They seem to comprise

a very self-sufficient economic community and are generally recognised as such by resi-

dents. The core household survey finds that some 40 per cent of black residents and 60 per

cent of minorities perceive the foreigners as offering real value for money as employees,

and they are also generally seen as self-sufficient, cooperative and trouble-free employees

who don’t create hassles for employers.

The attitudes of South Africans towards immigrants in the world of work are clearly far

from simple. The kinds of jobs and business activities that foreigners are perceived to per-

form so well are broadly the same jobs and opportunities that some Johannesburg resi-

dents consider foreigners to be taking away from the South African unemployed, without

thinking about the capacity or willingness of local people to perform these tasks at the

levels of remuneration available.

It is also important to note that minorities may be more worried about the jobs that immi-

grants take away from black South Africans than worried about their own jobs (a marked

feature of white attitudes to foreigners in CDE’s Witbank survey).

The sensitive issue of willingness to work at given remuneration levels tended to emerge

in the informal comments that many residents gave on how they distinguish foreigners,

rather than in formal responses. Black respondents in the household survey, and some-

times minorities as well, said that foreigners perform work that South Africans prefer to

avoid, even if unemployed. There is some hypocrisy in the notion that foreigners take jobs

and opportunities away from South Africans, because often the foreigners do not aspire

to, or initially aspire to, the wage and income levels that have become norms for most

local people. These norms are very close to the unionised wage levels in industry and

commerce, and the foreigners strategically pitch their expectations below these levels.

Acceptance and integration

How do the immigrants themselves experience the reactions of people in Johannesburg?

Nearly three-quarters of the foreigners are very or fairly happy in Johannesburg. However,

some 60 per cent of them experience significant hostility or negative treatment in Johan-

nesburg; even among those who are happy in Johannesburg, some 50 per cent experience

hostility. Men rather than women bear the brunt of hostility. The inner-city areas are least

39July 2008 |

The people who

bear the brunt of

xenophobia are those

that dress, look and

speak differently from

any South African

ethnic category


hostile to foreigners, mainly because of their cosmopolitan composition. People from the

BLNS countries, Europeans and East Africans experience least hostility, while Nigerians,

West Africans, people from the East and Mozambicans face substantial hostility.

Two main factors condition the foreigners’ experience and happiness in Johannesburg.

First of these is the attitude and manner of police and officials. We are realistic and accept

that this is a sensitive and disputed feature of the formal interface between foreigners and

host countries in many parts of the world.

The second is whether or not they are accepted as a South African or at least as part of

the South African community. As many as 56 per cent of the foreigners interviewed are

accepted as South Africans or as part of the community, which operates as a very effec-

tive neutraliser of what would otherwise be very damaging xenophobia. Therefore, the

ability to blend into the local community is the critical factor in the lives of Johannesburg

foreigners. It should be noted that some of the happy foreigners are well accepted by ordi-

nary residents, but experience hostility from officials and police, which accounts for the

apparent contradiction between high levels of hostility experienced and general satisfac-

tion with life in the city.

The foreigners that are most easily integrated, in order of acceptance, are Europeans, peo-

ple from the BLNS countries, Malawians, and Zambians. Trailing them, due to their large

numbers and great poverty, are Zimbabweans.

The crucial factors in becoming accepted as a South African include cultural compatibil-

ity, the ability to pass oneself off as belonging to one of the South African ethnic groups,

the ability to speak one of the core South African languages, and becoming part of South

African social networks. Most of the foreigners have a significant number of South African

friends.

The people who bear the brunt of xenophobia are those that dress, look and speak differ-

ently from any South African ethnic category – Nigerians and other West Africans, Soma-

lis, and, to a lesser extent, Mozambicans. They are either unhappy or have to establish

tight and segregated niche communities, which provide emotional security but do not

solve their problems with xenophobia.

How serious is xenophobia in Johannesburg? The short answer is that it would be utterly

crippling for foreigners were it not for the fact that so many foreigners have ethno-linguis-

tic and cultural characteristics that are fairly closely related to the ethnic categories that

exist in South Africa. South Africa has varieties of European, Nguni, Sotho, South Asian

and East Indian cultural templates, and the majority of the immigrants in Johannesburg,

white and black, share some or other of these very basic ethnic characteristics. In short –

xenophobia is tolerable for most of the immigrants but penalises some immigrant minori-

ties quite mercilessly.

Despite the fact that the foreigners have all the appearance of a fairly settled community,

some 75 per cent say that they intend to return to their countries of origin to retire. This

may be a nostalgic dream for many or most of them. We note from the findings that some

40 per cent never visit their home countries or do so less than once a year. We should

accept that, by the time they need to retire, far fewer than 75 per cent will actually return

home.



Generally the officials

lean significantly

towards negative

views on immigration

Key perceptions of local officials and business organisations

Local officials

The 32 local officials interviewed were all very senior in the national and metropolitan

police services, planning divisions of local government, regional immigration, and wel-

fare administration.

Key points in officials’ responses were:

Numbers

The key perception of respondents in this group was that the in-migration of foreigners

to Johannesburg is out of control. A large majority (84 per cent) of them observe that the

numbers of semi-legal or non-legal immigrants are increasing apace at present, and a

further 13 per cent see a ‘steady’ increase.

However, the most arresting indication of this belief in the failure of immigration control

is seen in the officials’ estimates of the numbers of foreigners in the city. Their estimates

average out at 2,5 million, of which 86 per cent are believed to be non-legal in status, with

a further 7 per cent semi-legal in the sense of being ‘corruptly legalised’ through the pay-

ment of bribes for falsified documents. Researchers may dismiss these estimates as indi-

cations of how ill-informed and poorly trained in numeracy even quite senior bureaucrats

are, but that misses the point. If we have officials who are close to the action on the streets,

and if they think that 2,5 million of Johannesburg’s official population of some 3,9 million

consists of (overwhelmingly irregular) foreigners, the estimates have disturbing implica-

tions for the morale of people administering the city.

This image of a process out of control inevitably produces feelings of helplessness or des-

peration among the officials who have to cope with its effects. The unreality of their own

estimates may be easily exposed, but they are fully aware that the authorities do not have

the faintest idea of how many foreigners there are or how many are coming in; at best

the authorities know how many people are deported, which our respondents on average

placed at just over 100 000 per year, and that knowledge is far from reassuring.

Generally the officials lean significantly towards negative views on immigration and its

impact in Johannesburg, although their views are certainly not one-sided or one-dimen-

sional, perhaps surprisingly in view of their perceptions of numbers. Broadly, only 12 per

cent of the officials see immigrants as an advantage for Johannesburg, and 55 per cent see

them as a liability with the rest (32 per cent) giving qualified answers.

Employment and roles in the economy

When specifically asked about the contributions of immigrants, many of the officials con-

cede that they make a significant contribution to small-business activity and are obvi-

ously ‘investing’ in the city in small but pervasive ways. They often admire the skills and

work ethic among the immigrants and concede that the immigrants are remarkably self-

sufficient as individuals.

Many of the officials perceived considerable occupational specialisation among foreign-

ers, with 71 per cent observing that foreigners mainly do particular jobs that locals do not

occupy. This in itself indicates that foreigners have a significant role in the economy.

41July 2008 |

Many officials observe

that foreigners mainly

do jobs that locals

do not occupy


The officials are very clear on why they think that people employ foreigners, and their

views tend to be cynical. The reasons they perceive are:

The labour is cheap and exploitable and some work mainly to get accommodation.•

Labour laws and union protection are circumvented.•

People pity them.•

People do not realise that the foreigners are undocumented.•

The foreigners have low aspirations and will do work that South Africans shun.•

The foreigners are compromised by their situation and hence are very cooperative or •

passive.

At the same time, however, almost two-thirds of the officials also admit that foreigners are

hardworking, determined, and productive.

Pressure on services and resources

The officials believe that all services, facilities, basic resources and agencies operating at

community level are critically oversubscribed or overloaded, and accretions of new peo-

ple of any kind will inevitably place critical stresses on the local urban system.

More specifically, the officials named significant stresses and pressures as follows:

Available housing space•

Law enforcement and the criminal justice system•

Capacity of the infrastructure•

Availability of employment•

Services, including welfare and education•

Health services•

Control of street trading, begging and the condition of the streets in general•

Immigration control through bribery, illegal marriages and costs of repeated •

deportations

Local revenue through non-payment of rates and taxes.•

In the officials’ view, these economic costs, plus the knock-on effects on the investment

image of the city, cost the city more than the economic benefits that the foreigners bring.

A particular concern among the officials is the amount of time and resources that have

to be spent in deporting foreigners. Over six out of ten are convinced that foreigners

deported at great cost return very soon. Seven out of ten are aware of violence and con-

flicts between locals and foreigners that have to be controlled, and seven out of ten there-

fore see the police as spending a disproportionate amount of their time on managing the

consequences of a system that has collapsed.

Crime

Johannesburg already has to cope with high levels of crime, drug use and sexual exploita-

tion and violence, and it is a widespread public perception that foreigners worsen the

situation quite significantly, or at least undermine the effectiveness of law enforcement by

tying up police personnel in immigrant control. However, views on the criminal propensi-

ties of foreigners vary among the security personnel interviewed. The following views sum

up their perceptions:

Foreigners more likely than locals to be involved in crime: 50 per cent•



Views on the criminal

propensities of

foreigners vary

among the security

personnel interviewed

Foreigners and locals equally involved: 43 per cent•

Foreigners less likely to be involved: 7 per cent.•

Hence about half the officials believe that foreigners are no worse than locals in their

criminal propensities. Among those who perceive the foreigners to be more vulnerable to

criminal involvement, there are some who add that the circumstances of many foreign-

ers force them into crime. Another specific amplifier of the crime issue among foreigners

is that they are so often unrecorded immigrants and hence are untraceable in criminal

investigations. On balance the views of the officials are reasonably rational – the police

officials are not crudely xenophobic by any means, and they motivated their views very

plausibly.

The officials are not fundamentally opposed to immigration as such. We asked three dif-

ferent questions about skills needs, the contributions of ‘legal’ immigrants and the kind

of people Johannesburg needs more of. There was broad consensus about the following

desirable immigrants:

Engineers•

Other professionals•

Teachers and educators•

Entrepreneurs and cross-border traders•

Personnel for the construction industry•

Experienced people to work on important projects•

Artisans and technicians•

Manufacturing industrialists•

IT specialists•

A range of other, similarly useful, skills.•

The balance sheet of immigration: officials’ views

The officials conceded the past contributions of foreign artists, industrialists, craftspeople,

chefs, sports coaches, medical experts, entertainers and the like. Many of their answers

emphasised the value of immigrant contact with the global economy, the advantages of

cultural diversity, and the link between immigration and foreign investment. Some had

lower expectations of useful immigrants, asking only that they should pay their taxes.

Hence the officials are far from being sullenly parochial in their cynicism about current

inflows. They do not appear to be xenophobic.

Nor do the officials think that Johannesburg residents are xenophobic. Some eight of

them were adamant that there is no xenophobia in Johannesburg; more felt that hostility

towards foreigners is mild under the circumstances, but three of them felt that there is

unnecessary aggression among law enforcement personnel. In talking about xenopho-

bia the officials were at pains to point out that it is difficult for people to feel warm and

welcoming about people who break the law, pay no taxes and who seem to be flooding

in at an increasing pace. Generally they argued that what appeared to be xenophobia was

mainly a response among people who are themselves struggling to survive.

Finally, when asked about their prescriptions for policy, most of the officials argued for

better immigration controls, including improved border surveillance, more capacity and

effectiveness in the Department of Home Affairs and, the deployment of special task

groups in immigration control. They particularly wanted the authorities to establish the

43July 2008 |

Most of the officials

argued for better

immigration controls

and a more effective

Department of

Home Affairs


true numbers of immigrants as a basis for strategy. Some recommended more cooperation

between the sending countries and South Africa, including control over irregular emigra-

tion in such countries. One person said that the government should stop the hypocritical

practice of employing prominent foreigners in high paying positions, while pretending to

control immigration on the ground.

Understandably, most of their prescriptions focused on better control of immigration, but

some added that if controls became too strict they would be violated. They often agreed

that arresting and deporting undocumented foreigners was a waste of time and resources.

At least a third accepted that scarce skills should be welcomed into the country, and their

replies made it clear that such skills included entrepreneurial talents.

Business organisations

The views summarised here were expressed in interviews with senior people in all 45

significant business organisations operating in Gauteng, including major chambers of

industry associated with foreign countries. Rather than a sample, then, this is a repre-

sentative census of the viewpoints of organised business on an issue that is close to their

core interests. Average membership of these bodies in Johannesburg alone is 11 300 busi-

nesses, with around 20 per cent of the membership being composed of small businesses

and 40 per cent of large corporations.

Key points in business responses were:

The pattern of distinct unease expressed in many people’s perceptions of immigration to

South Africa today was mirrored in the responses of business organisations, which con-

firmed the perceptions of loss of control already recorded.

How problematic are large numbers of undocumented foreigners?

A very serious problem: 49 per cent•

A fairly serious problem: 27 per cent•

Both a problem and a benefit: 16 per cent•

A benefit rather than a problem: 9 per cent.•

As with the officials, the balance of views is very negative. The perceptions of the business

organisations are also that the problem is escalating.

Growth or decline in the numbers of undocumented foreigners:

Numbers growing rapidly: 53 per cent•

Numbers growing steadily: 44 per cent•

Declining: 2 per cent.•

Numbers

When asked to estimate numbers, business respondents who were prepared to hazard

an estimate (32 out of the 45) put the total number of immigrants in Johannesburg at

2 766 000 on average. Like the officials (whose estimates they slightly exceed) they too see

this as a problem out of control. Around 70 per cent of the business respondents feel that

the common estimate of the numbers of non-legal foreigners in South Africa made by the

police and the government, namely some 2,5 to 3 million, is an underestimate, and less

than one-third think that the estimate is about right.



Business respondents

feel that business

start-ups by

immigrants add more

than they subtract

from opportunities

for local people

Skills and the economy

This does not mean to say that they do not see benefits in the presence of foreigners. For

example, when we asked them whether they see immigrants contributing to small busi-

ness start-ups, 20 per cent said that the contribution was considerable, and a further 44

per cent felt that it happened to some extent.

Where it happens, the business respondents feel that business start-ups by immigrants

add more than they subtract from opportunities for local people. Only 24 per cent of them

feel that foreigners reduce the business opportunities for locals, and 76 per cent feel that

they add value, with initiatives that would not happen without them. The organisations

add that there are genuine shortages of work experience and practical skills, even at

‘unskilled’ levels that foreigners help to address. Hence even if foreigners are taking away

certain opportunities from locals, they are very often adding value in the process. Exam-

ples of this given by the business organisations include the following detailed points:

Foreigners create jobs even at low levels of skill and complexity.•

There are many specialised skills in SA that can be traced back to foreign inputs.•

Foreigners provide contact with Africa that we would otherwise not have.•

Foreigners introduce styles of dress and products that attract tourists.•

The introduction of foreign food adds value.•

They create role models of a work ethic.•

Their communication eases the fear of South Africa that exists in many parts of •

Africa.

Foreigners are known to take over failing businesses and make a success of them, eg •

Somalis.

They provide low-cost maintenance and repair services.•

They have a large impact on the recycling market.•

Whatever the benefits, however, the business organisations are aware of pervasive prob-

lems in the scope and activity of foreign migrants in Johannesburg. They are painfully

aware of the widespread feeling that the immigrants take work and opportunities away

from local jobseekers or street traders. Part of the problem lies in the responses of the

unions, but the tensions would exist anyway. The responses of the business organisations

remind us once again that as long as the government cannot adequately control and steer

the process of foreign in-migration, the benefits of immigration, even at the levels of skill

at which it is occurring, could be lost in the tension and conflict created by a process that

is widely perceived to be out of control.

One of their central objectives is to make representations to government in respect of their

business needs and interests. In general the response that they get from government is as

follows:

Helpful and responsive: 44 per cent•

Not helpful: 24 per cent•

Qualified reactions of various kinds: 31 per cent.•

In the light of the importance of the interests that they represent, the 24 per cent who

experience an unhelpful reaction is rather high. Among the issues which cause problems

in representations to government, and on which government is least helpful, are labour

and employment issues; education and immigration issues are particularly prominent,

with eight organisations making special mention of difficulties in engaging the DHA.

45July 2008 |

Some 40 per cent

of the organisations

observe that their

members frequently

have difficulties in

recruiting suitably

skilled staff


Representations on immigration issues and work permits are not a major ongoing func-

tion of any of the business organisations, but are significant functions of the organisations

nevertheless.

Some 40 per cent of the organisations observe that their members frequently have dif-

ficulties in recruiting suitably skilled staff; a further 48 per cent have fairly frequent diffi-

culties. This is a sobering finding. Specific skills shortages that the business organisations

mentioned, in order of the number of mentions, were:

Management and project management skills•

Artisan skills•

Other technical skills (excluding IT)•

Financial skills•

IT skills•

Engineering•

Sales and marketing•

General administrative skills.•

There were many others at lower levels of mention.

We need to understand these skills constraints in the light of the experience of the busi-

ness organisations with the SETA-based training scheme funded by the business levy-

based National Training Fund.

Reactions to the sector based (SETA) learnerships:

Generally very useful: 25 per cent•

Moderately useful: 21 per cent•

Some useful, others not: 27 per cent•

Most not really useful: 16 per cent•

Not useful at all: 11 per cent.•

These reactions indicate that the SETA-based training is making a useful contribution but,

with around 27 per cent of the organisations expressing reservations, there are clearly

gaps in the production of new skills. The objectives of the SETAs are to produce skills in

standard categories, however, and they were never intended to be a vehicle for producing

specialised skills.

Specific problems with SETA-based training included the following major issues:

Positive, useful contribution to skills shortage/help trainees to get work experience •

and become marketable, etc: 16 mentions

Keeps industry up to date in techniques: 3 mentions•

No other alternatives: 2•

Sound concept but inefficiently implemented: 13 mentions•

Insufficient incentives for employers in relation to time and effort in claiming back: •

19

Poor management/disorganised: 20•

Poor communication with employers: 8•

Bureaucratic: 5•

Training not practical, relevant or focused enough: 17•

Complex interaction required to derive benefit: 7•

Technical mismatches in training: 2•

Long delays in accreditation and placements: 4•



Two-thirds of business

organisations represent

members who need

to access and recruit

foreign skills

Training at too low a level: 2•

SETAs have skills constraints: 2•

Learners are confused about whether employed or not: 2•

Employer-based training more effective: 2•

As a consequence of both the gaps and the levels of training offered by SETAs, it is not

surprising that some 30 per cent of the organisations observe a great need among their

members to recruit foreigners, and a further 36 per cent have a moderate need for foreign

recruitment. Hence some two-thirds of the organisations represent members who need

to access and recruit foreign skills.

The skills that the organisations consider to be in short supply in South Africa, and that are

very likely to be found in foreign recruitment include, in order of mention:

Artisans (non-construction-based)•

Engineers•

Other technical skills of various kinds (non-IT)•

Management and project management skills•

Financial skills•

Various applied professions•

IT skills•

Construction technicians and highly skilled artisans•

Health professionals•

Entrepreneurs•

Planning and strategy skills•

Catering skills•

Marketing skills•

A range of others, including R and D, art and design.•

Ironically, seven organisations also claimed that their members were short of cost-

effective, efficient and productive unskilled labour. This small finding is but one of many

indications that South African unskilled labour may have priced itself out of the market.

Another point made was that increased diversity in the supply of labour would do much

to improve efficiency.

Clearly then, the organisations representing South African industry consider that foreign

skills have a major role to play in the further development of the economy.

Against this background we asked them about experiences with the Department of Home

Affairs. They responded as follows:

Very helpful: 4 per cent•

Fairly helpful: 22 per cent•

Not at all helpful: 51 per cent•

Do not use them/have not had to use them: 22 per cent.•

More detailed accounts of their experiences with the DHA can be broken down as

follows:

Bureaucratic/rigid/petty/sluggish: 28 mentions•

Various problems with visas and work permits: 29 mentions•

General problems of communication that are encountered with all departments, not •

only DHA: 17 mentions

47July 2008 |

The recruitment

of foreign skills is

obviously highly

cost-effective for

an economy, but

has wider political

and socio-economic

implications


Experiences of corrupt officials: 3 mentions•

Evidence of breakdown of systems of control and the need for tamper-proof docu-•

ments: 3 mentions

Unsympathetic attitude: 2 mentions.•

The recruitment of foreign skills is obviously highly cost-effective for an economy, but in

South Africa it has wider political and socio-economic implications, as the earlier parts of

this report have indicated. Part of the complexity of the issue is that we are said to have a

problem of xenophobia. Hence we asked the organisations how, in their experience, local

staff react to and treat foreign recruits, and the following pattern emerged:

Acceptance at higher levels of skill or in specialised skills areas: 18 mentions•

Negatively/with resistance, particularly at lower skills levels and among unionised •

blue collar workers: 23 mentions

Negatively because of the belief that SA has all the skills it needs: 11•

Dislike of competition from foreigners: 11•

Evidence of racism in reaction to foreigners: 5•

Resignation/reluctant acceptance: 3•

Other and qualified answers.•

We also asked the organisations about the problems that foreign recruits have in adjusting

to South Africa. The effects of high crime levels on foreigners were mentioned above all

other effects. Mention was also made of problems in finding suitable accommodation and

the lack of adequate public transport. Social integration into communities with a different

culture was mentioned, as was adjusting to a different pace of work and occupational cul-

ture. Entrenched antagonism and xenophobia emerged as a middle-level problem; nine

of the organisations said that they had not noticed any particular problems.

In concluding this section, we must observe that there is considerable negativity about the

foreign presence among both key officials and business organisations. They display the

underlying cause of this negative response in their estimates of the numbers of foreigners

in the city. Their astoundingly high estimates are clearly unrealistic, but they are important

in that they reflect the perception that the present state of immigration control is in chaos.

Estimates of more than two-and-a-half million foreigners on average tell us that they see

the government as having lost control completely. What they express through dramati-

cally high estimates, ordinary residents are expressing in their apparent xenophobia.

Summary of results

The method of estimating foreigners within households adjacent to sampled dwell-•

ings, with safeguards and corrections, is a more useful method of getting closer to the

real numbers of unrecorded foreigners than any other approach. However like other

methods of estimating foreign presence, it probably yields conservative results.

The numbers of foreigners in Johannesburg at the time of the study (2006) are esti-•

mated to have been over 500 000, or 14,1 per cent of the estimated adult population.

Due to the under-representation of identified types of foreigners, however, in reality

the number could be somewhat higher, most likely nearer 600 000 to 700 000 today.



There is much less

direct job competition

between foreigners

and locals than

people imagine

This however is not a firm finding of this research, and we should adhere to our esti-

mate of 500 000 to 550 000.

The method has other advantages, however, since it provides a roughly representa-•

tive selection of most kinds of foreigners interviewed who live in the residential fabric,

excluding servants’ quarters.

The method is useful in empirical research, but is not recommended for large mass or •

commercial surveys because of the need for high-quality face-to-face interviews with

qualitative components.

The foreigners in Johannesburg have significant skills to offer, although these are more •

often than not experience-based rather than certificated skills.

On average the foreigners are not highly educated, although the average level of edu-•

cation (incomplete high school) is often of a higher standard than equivalent South

African schooling.

The foreigners are more than twice as likely to be self-employed and self-sufficient as •

local adult residents. Their level of unemployment, which occurs mainly among very

recent immigrants, is also significantly lower than the South African equivalent.

The immigrants employ almost half their total numbers, almost half again being South •

African employees.

Up to one-third of the immigrants take paid employment that unemployed South Afri-•

cans say they would like to have, but the immigrants are grateful for the opportunity,

whereas the jobs would be regarded as inferior or underpaid by local standards. In

reality, there is much less direct job competition between foreigners and locals than

the estimate of up to one-third would suggest.

Although they claim to intend returning to their countries of origin, we should accept •

that very substantial proportions of immigrants are likely to remain in South Africa for

the rest of their productive lives, and many will even retire here.

Most immigrants integrate well into the South African community, with many South •

African friends, and nearly 60 per cent feel accepted as part of the local community.

Most of them are happy with life in Johannesburg, rather more so than local people.

Xenophobia is pervasive in a broad sense, and is associated with an accumulation of •

a large variety of specific negative attitudes rather than a consolidated mindset. It is

caused less by competition for jobs and resources than we expected, and at least as

much by personality attributes and ethnic loyalties among the local people. However,

it is not the result of colour-based racism on the part of black or white South Africans.

The accumulated specific hostilities than make up the overall picture of ‘xenophobia’ •

is important in that it does cause unhappiness among foreigners. The reactions of offi-

cials and police are a major cause of the unhappiness.

The key potential effects of hostility are muted, however, by the fact that a majority of •

the foreigners interviewed feel accepted in the local community. On balance, there-

fore, xenophobia is far from being a crippling problem for immigrants.

The national origins of immigrants are an important intervening variable, however, •

and the people from more distant countries in Africa, excluding English-speaking East

Africa, are exposed to considerable hostility and stereotyping. The reasons are pre-

dominantly cultural, although they are expressed as an assumed proclivity for crimi-

nal behaviour.

The hostility to foreigners does not extend to a reluctance to employ or trade with for-•

eigners. In fact their utility as employees and people in business is generously recog-

nised by South Africans, by businesspeople, and by many key functionaries.

49July 2008 |

The hostility to

foreigners does not

extend to a reluctance

to employ or trade

with foreigners


The cultural diversity that immigrants bring is not disruptive in general, and some •

of the black immigrants who are culturally quite distinctive have broken through the

informal residential segregation of a South African city. However, less well-educated

minorities, and even better educated black South Africans, find it difficult to enjoy

contact with people from culturally remote areas of Africa. Our interpretation of the

patterns, however, is that the additional diversity that immigrants bring is helping to

blur the edges of inter-ethnic hostility among South Africans.

Concluding remarks

CDE’s field research was undertaken in the second half of 2006 at a time when concern was

mounting at escalating numbers of irregular migrants in South Africa. However this anxi-

ety was not backed by any hard data on numbers. Xenophobic attacks were not unknown

in the country at that time, but they were sporadic, isolated, targeted (at particular nation-

alities, including Somali shopkeepers) and appeared to follow local dynamics.

CDE’s goals in undertaking the research were ambitious, but limited. They were ambi-

tious because enumerating undocumented migrants is very difficult and little fieldwork

had been done in South Africa – on any scale at least – to provide the hard numbers that

sound public debate and credible policies require.

The goals were limited in that they were confined to one metropolitan area and to the

specific goals of counting foreigners, constructing a profile of them, and reporting on the

perceptions and attitudes of Johannesburg residents to their presence and impact.

From this perspective, what did we find out?

NumbersWe believe that we are reporting a more credible number of foreigners in Johannesburg

in the figure of upwards of half a million or slightly more than 14 per cent of the estimated

population of 3,9 million. This is less than the popular imagination, and much less than

the greatly exaggerated figures volunteered by the highly placed officials and business

organisations that we interviewed. We believe that this figure is more credible because it

is based on actual field research.

However, we are too conscious of the difficulties of research in this field and the limita-

tions of our own methodology – which we have been careful to point out – to claim that

this is an authoritative figure. Complete authority will remain elusive on this subject. It is

worth recapping some of the caveats which we believe apply.

We make no finding on the specific numbers of irregular migrants, nor on those who •

have genuine or forged documents (ID books, refugee permits, work permits, study

permits, etc). Our interviewers strongly suspected that the problem of forged docu-

ments was a large one, but this simply could not be quantified.

We believe that our findings are likely to be an underestimate, partly because of the •

problem of ‘reach’ – for instance in the anonymity of servants’ quarters in the suburbs.

It is also possible that there has been a sudden and dramatic spike in numbers since



CDE’s methodology

and findings provide

the basis for a more

reasoned and informed

public debate on

immigrant numbers

late 2006. We believe that even if this were so, it would not account for millions more

immigrants.

However, even if our figures are an underestimate, some of the figures that have become

common currency challenge logic and simple arithmetic. If there are 9,84 million irregu-

lar immigrants in the country, South Africa is hosting an underground population two-

and-a-half times the population of Johannesburg, as estimated in the Community Survey

of 2007. If there are 3 million Zimbabweans in South Africa, this presence amounts to

more than three-quarters the population of Johannesburg.

At the very least, our methodology and our figures provide the basis for a more reasoned

and informed public debate on numbers, and a challenge to anyone who believes in

the ‘tidal wave’ of immigrants to prove us wrong with alternative estimates based on

fieldwork.

At most we would hope that our figures might give confidence to policy-makers that man-

agement of these – still admittedly large – numbers is feasible, and that policies that will

restore public confidence are achievable.

Impact and attitudesCDE’s research findings in both the pilot survey of Witbank and the Johannesburg survey

found ample evidence of hostility to foreigners. However in both instances, there was also

evidence – mainly from the foreigners themselves – that a kind of modus vivendi existed,

allowing South Africans and foreigners to get along tolerably well. Clearly, however, since

the research was completed, some sort of tipping point was reached which in certain

specific areas led to a chain reaction of violent hostility to develop and break down this

mutual tolerance (see box: This thing has always been there . . ., facing page).

‘This thing has always been there but it was a light thing’

‘We were against these people from the onset: that’s when the term like makwerekwere

(derogatory term for foreigners) came about, we were against them in a light manner but now

people are getting angry that is why they beat them up, their numbers are growing and some

have babies this side it’s as if this is their hometown; this violence happened because people are

getting angry, this thing has always been there but it wasn’t as strong as it is now. We never said

we are happy to live with them but it was a light thing so people resorted to violence because of

the realisation that the situation was getting serious.’

Focus group respondent, HSRC Xenophobia study, June 200822

CDE 2008

A combination of several factors has probably brought us to this tipping point:

The escalation of Zimbabwe’s political and economic crisis following the 29 March •

2008 elections drew attention to the refugee numbers that had been growing visibly

but uncounted for over a year.

The sense of a flow of refugees that was out of control was exacerbated by the absence •

of any display of leadership, sense of urgency or organised response by the South Afri-

can government (see box: ‘Just live with it’ below).

51July 2008 |

Investigation of the

causes of violence

should be an

urgent priority


The sharp downturn in the economy and sharp rise in the cost of living that imme-•

diately preceded the xenophobic attacks produced a generalised climate of unease,

tension and protest – foreigners became one among several foci.

Many or most of the focal points of violence saw looting of prominent immigrant •

shops. With the retreat of township retail commerce into shopping centres or mini-

spaza shops, the prospect of looting these immigrant-owned retail outlets was an

incentive for more general unrest.

These factors point to a chain of social, economic and political causation much more com-

plex than viewing the violence as the outcome of a spasm of ‘xenophobic violence’. Inves-

tigation of the causes of violence should be an urgent priority and preferably it should be

carried out by an independent, well-resourced and high-profile judicial commission.

‘Just live with it’

‘As for Zimbabweans who enter South Africa legally, well, they enter South Africa legally and

there wouldn’t be any need to do anything about that, but as to this other influx of illegal people,

I personally think it’s something that we have to live with … You can’t put a Great Wall of China

between South Africa and Zimbabwe to stop people walking across…’ (emphasis added)

President Thabo Mbeki in the National Assembly, 17 May 2007

CDE 2008

On the other hand, while it is important to bear causal factors like these in mind, it would

be wrong to treat hostility towards foreigners as merely some sort of displacement behav-

iour by people driven to strike out in anger by the poor conditions of their lives.

In this regard it is worth recalling some of the conclusions of the CDE Witbank report:

Large majorities of every racial group among Witbank residents expressed

uncompromisingly negative views and attitudes towards immigrants . . .

Despite this:

. . . our interviews with immigrants showed that much of the hostility expressed

is defensive rhetoric, and that by and large, Witbank treats its foreigners rather

well. Nearly 90 per cent of the immigrants felt good about the way they were

treated, more than two thirds like Witbank and its people and most of them said

that they would like to stay . . .

Although on this basis we concluded that ‘host’ citizens had at least a grudging tolerance

of foreigners, we issued a warning:

Negative attitudes by no means always translate into hostile acts . . . We need

to know much more about the triggers that could turn negative attitudes into

hostile action.

In our much larger Johannesburg study we found broadly similar effects, with the excep-

tion that Johannesburg residents were less, rather than more, likely than their Witbank

counterparts to express hostility to immigrants, while immigrants in Johannesburg were

much more conscious of ill-treatment from officials and police than their counterparts in

Witbank.



There is no reason to

panic about immigrant

numbers, but there

are no grounds for

complacency either

It is unlikely that large majorities of immigrants in our pilot and main surveys independ-

ently volunteered false information about the degree of acceptance that they encountered

and integration that they felt. As a result it would seem that there was a modus vivendi and

that it has broken down, at least temporarily and in places.

Managing migration in the national interestCDE’s findings on immigrant numbers and impact may show that there is no reason to

panic about immigrant numbers, but there are no grounds for complacency either. If we

are to restore, never mind improve on, the precarious tolerance that broke down under the

pressures we have described, the direction in which our analysis clearly points is the need

for a comprehensive and effective strategy to deal with immigration issues. This needs to

be coupled with clear and bold leadership to build public confidence in the government’s

ability to deal with the realities of migration.

What are those realities? As the country – and especially Johannesburg – counts the cost of

the May 2008 violence, we would point to three realities that will shape future challenges

of immigration policy.

The first is that continuing and probably increased immigration is inevitable. South Africa

– Africa’s richest economy – has long and porous borders at the bottom of a poor and

conflict-prone continent. Increased migration pressures from climate change to improved

transportation to political conflict will push out-migration from sending countries, while

the performance of South Africa’s economy will act as a magnet for those with the get-up-

and-go to leave their own countries and try for a better life elsewhere.

The second is that most immigrants, at all levels – from the street trader to the boardroom

– make a contribution to South Africa’s economy, sometimes against daunting odds. We

need to understand, manage and capitalise on that contribution much better. But we will

only be able to do this if we are able to manage flows of people across our borders more

rationally and efficiently, that is in ways that give South Africans confidence that these

flows of people make a positive contribution to the common good and that the country is

controlling its own borders.

This means, thirdly, that popular fears and misconceptions about immigration must be

dealt with. All over the world migration issues are disputed between elite concerns for

skilled labour needs and human rights on the one hand, and popular concerns about

preferential treatment over locals, unfair competition in the labour market, border control,

and bogus claims to asylum. When they show themselves here, these concerns are not the

products of blind prejudice or some national predisposition to xenophobia, but local ver-

sions of universal and understandable fears. None of them is as a simple as it looks and all

may be fanned by misinformation. But if immigration is to make the contribution it can to

our economic growth, then they have to be taken seriously and treated on their merits.

This report and the research on which it is based were not intended to produce a strat-

egy for addressing these realities. This aim belongs to a larger and more ambitious CDE

study which covers all aspects of immigration policy but with a focus on skills needs and

the economy. The forthcoming final report of this project will provide a comprehensive

overview of migration policy issues facing South Africa and include research-based rec-

ommendations for more effective policy making in the country’s approach to migration

management.

CDE’s findings on immigrant numbers and impact may show

that there is no reason to panic about immigrant numbers,

but there are no grounds for complacency either


Endnotes

Endnotes1 Forced Migration Studies Programme (FMSP), Wits University, Xenophobic violence: myths

and facts, media release, 22 May 2008.

2 Lawrence Schlemmer with MarkData, Between hostility, acceptance and hope: foreign

immigration in a medium-sized industrial town in South Africa, CDE, Johannesburg,

November 2005.

3 Lawrence Schlemmer with MarkData, Immigrants in Johannesburg: foreign immigration in

South Africa’s largest metropolis, CDE, August 2007.

4 The team was made up of Charles Simkins, Helen Suzman Professor of Political Economy at

Wits University; Prof Jackie Galpin, head of the School of Statistics and Actuarial Science at

Wits University; and H A Steenkamp of the Bureau of Market Research, University of Pretoria.

5 Statistics SA, Community Survey 2007, Municipal data on household surveys, report

number 03-01-22, pp xi–xv, http://www.statssa.gov.za/Publications/Report-03-01-21/

Report-03-01-212007.pdf (accessed on 21 May 2008).

6 National Treasury, media release, 3 October 2007, http://www.treasury.gov.za/legislation/

mfma/media_releases/mbi/01.%20Press%20Release%20-%202007%20MTREF%20-%20

04%20Sept%202007.pdf.

7 K Dunnell and others, Globalisation: what are the main statistical challenges? Economic and

Labour Market Review, 1:9, September 2007, pp 18–24. For reaction to the ONS report, see

Migration statistics challenge for ONS, Financial Times, 10 September 2007; and Chaos over

immigration count ‘is a threat to the economy’, Daily Mail, 10 September 2007.

8 Stats SA, Report-03-51-02 – Tourism, 2006, 12 March 2008, http://www.statssa.gov.za/

publications/statsdownload.asp?PPN=Report-03-51-02&SCH=4120 (accessed on 21 May

2008).

9 Stats SA, Report-03-51-02 – Tourism, 2006, 12 March 2008.

10 United Nations Population Division, http://www.un.org/esa/population/unpop.htm

(accessed on 7 August 2008).

11 United Nations Population Division; Stats SA, Report-03-51-02 – Tourism, 2006, 12 March

2008.

12 Fin24, More illegals set to flood SA, 23 November 2006.

13 United Association of South Africa (UASA), 4th Annual Employment Report, http://www.

uasa.co.za/reports/EmpReportNo4.pdf (accessed on 20 May 2008).

14 E Sisulu and others, The Zimbabwean community in South Africa, in S Buhlungu and others

(eds), State of the Nation: South Africa 2007, p 554.

15 Solidarity Peace Trust, No war in Zimbabwe: an account of the exodus of a nation’s people,

2004.

16 Loren Landau, Drowning in numbers, in Migration from Zimbabwe: numbers, needs and

policy options, CDE, Johannesburg, April 2008, pp 7–11.

17 Figures drawn from annual reports of the Department of Home Affairs: 2003–4 (p 3); 2004–5

(p 22); 2005–6 (p 22); 2006–7 (p 25).

18 South African Government Information Service, Home Affairs anti-corruption campaign

nabbed 15 immigration officials, http://www.info.gov.za/speeches/2007/0712211015`1001.

htm (accessed on 20 May 2008).

19 Department of Home Affairs, Annual Report 2006–7, p 75.

20 See Counting immigrants in cities across the globe, Migration Information Source, a project

of the Washington-based Migration Policy Institute (MPI). The MPI’s figures come from the

55July 2008 |

Globalization, Urbanization and Migration website of George Washington University, www.

gstudynet.org/gum.

21 The population census estimated net immigration to Gauteng between 1996 and 2001.

According to the Labour Force Survey of Stats SA, 2004, net immigration to the three Gauteng

metros for the period 1999–2004 was 250 850. See C Simkins, South African metropolitan

areas: economic growth and job creation, statistical synthesis report for CDE, June 2006.

22 Human Sciences Research Council, Citizenship, violence and xenophobia in South Africa:

perceptions from South African communities, June 2008.

CDE’s estimate places immigrant numbers well above any level

that could be an excuse for complacency, and our analysis clearly

points to the need for a comprehensive and effective strategy to

deal with immigration issues. This needs to be coupled with clear

and bold leadership to build public confidence in the government

and the city’s ability to deal with the realities of migration

Cover: A trader sells Zimbabwean handicrafts in Bruma, a suburb east of the

Johannesburg CBD. Henner Frankenfeld / PictureNET Africa

Designed and produced by Acumen Publishing Solutions, Johannesburg

Printed by The Bureau, Johannesburg

Previous publications in this series

South Africa’s ‘Door Knockers’: Young people and unemployment

in metropolitan South Africa (July 2008)

Under the radar: Pentecostalism in South Africa and its potential social and

economic role (March 2008)

The skills revolution: Are we making progress? (October 2007)

Skills, growth and migration policy: Overcoming the ‘fatal constraint’ (February

2007)

Can black economic empowerment drive new growth? (January 2007)

Private schooling for the poor? International research raises new questions for

South Africa (November 2005)

Growth and development in South Africa’s heartland: Silence, exit and voice in

the Free State (July 2005)

Labour-intensive public works: Towards providing employment for all South

Africans willing to work (April 2003)

BOARDL Dippenaar (chairman), A Bernstein (executive director), F Bam, E Bradley, C Coovadia,

A De Klerk, B Figaji, S Jonah, S Maseko, I Mkhabela, S Ndukwana, W Nkuhlu, S Ridley,

M Spicer, E van As, T van Kralingen

INTERNATIONAL ASSOCIATEProf Peter Berger


immigrants in johannesburg: estimating numbers and assessing impacts

Documents